Datasets
Models
Downloads: 1
Diff of
/datasets_csv/Preprocessing.ipynb
[000000]
..
[405115]
Switch to unified view
| a | b/datasets_csv/Preprocessing.ipynb | ||
|---|---|---|---|
| 1 | { |
||
| 2 | "cells": [ |
||
| 3 | { |
||
| 4 | "cell_type": "code", |
||
| 5 | "execution_count": 1, |
||
| 6 | "metadata": {}, |
||
| 7 | "outputs": [], |
||
| 8 | "source": [ |
||
| 9 | "import os\n", |
||
| 10 | "from os.path import join\n", |
||
| 11 | "\n", |
||
| 12 | "import pandas as pd\n", |
||
| 13 | "import numpy as np\n", |
||
| 14 | "\n", |
||
| 15 | "label_col = 'survival_months'\n", |
||
| 16 | "n_bins = 4\n", |
||
| 17 | "eps = 1e-6" |
||
| 18 | ] |
||
| 19 | }, |
||
| 20 | { |
||
| 21 | "cell_type": "code", |
||
| 22 | "execution_count": 2, |
||
| 23 | "metadata": {}, |
||
| 24 | "outputs": [], |
||
| 25 | "source": [ |
||
| 26 | "def add_bins(slide_data):\n", |
||
| 27 | " assert 'case_id' in slide_data.columns and 'censorship' in slide_data.columns\n", |
||
| 28 | " \n", |
||
| 29 | " patients_df = slide_data.drop_duplicates(['case_id']).copy()\n", |
||
| 30 | " uncensored_df = patients_df[patients_df['censorship'] < 1]\n", |
||
| 31 | " disc_labels, q_bins = pd.qcut(uncensored_df[label_col], q=n_bins, retbins=True, labels=False)\n", |
||
| 32 | " q_bins[-1] = slide_data[label_col].max() + eps\n", |
||
| 33 | " q_bins[0] = slide_data[label_col].min() - eps\n", |
||
| 34 | "\n", |
||
| 35 | " disc_labels, q_bins = pd.cut(patients_df[label_col], bins=q_bins, retbins=True, labels=False, right=False, include_lowest=True)\n", |
||
| 36 | " patients_df.insert(2, 'label', disc_labels.values.astype(int))\n", |
||
| 37 | "\n", |
||
| 38 | " patient_dict = {}\n", |
||
| 39 | " slide_data = slide_data.set_index('case_id')\n", |
||
| 40 | " for patient in patients_df['case_id']:\n", |
||
| 41 | " slide_ids = slide_data.loc[patient, 'slide_id']\n", |
||
| 42 | " if isinstance(slide_ids, str):\n", |
||
| 43 | " slide_ids = np.array(slide_ids).reshape(-1)\n", |
||
| 44 | " else:\n", |
||
| 45 | " slide_ids = slide_ids.values\n", |
||
| 46 | " patient_dict.update({patient:slide_ids})\n", |
||
| 47 | " \n", |
||
| 48 | " return q_bins, patient_dict, patients_df" |
||
| 49 | ] |
||
| 50 | }, |
||
| 51 | { |
||
| 52 | "cell_type": "code", |
||
| 53 | "execution_count": 3, |
||
| 54 | "metadata": {}, |
||
| 55 | "outputs": [], |
||
| 56 | "source": [ |
||
| 57 | "slide_data = pd.read_csv('./tcga_gbmlgg_all_clean.csv.zip', compression='zip', header=0, index_col=0, sep=',', low_memory=False)\n", |
||
| 58 | "\n", |
||
| 59 | "n_bins = 4\n", |
||
| 60 | "eps = 1e-6\n", |
||
| 61 | "\n", |
||
| 62 | "### Asserts that 'case_id' is a column, not an index.\n", |
||
| 63 | "if 'case_id' not in slide_data:\n", |
||
| 64 | " slide_data.index = slide_data.index.str[:12]\n", |
||
| 65 | " slide_data['case_id'] = slide_data.index\n", |
||
| 66 | " slide_data = slide_data.reset_index(drop=True)\n", |
||
| 67 | "\n", |
||
| 68 | "q_bins, patients_dict, slide_data = add_bins(slide_data)\n", |
||
| 69 | "\n", |
||
| 70 | "slide_data.reset_index(drop=True, inplace=True)\n", |
||
| 71 | "slide_data = slide_data.assign(slide_id=slide_data['case_id'])\n", |
||
| 72 | "\n", |
||
| 73 | "label_dict = {}\n", |
||
| 74 | "key_count = 0\n", |
||
| 75 | "for i in range(len(q_bins)-1):\n", |
||
| 76 | " for c in [0, 1]:\n", |
||
| 77 | " label_dict.update({(i, c):key_count})\n", |
||
| 78 | " key_count+=1\n", |
||
| 79 | "\n", |
||
| 80 | "for i in slide_data.index:\n", |
||
| 81 | " key = slide_data.loc[i, 'label']\n", |
||
| 82 | " slide_data.at[i, 'disc_label'] = key\n", |
||
| 83 | " censorship = slide_data.loc[i, 'censorship']\n", |
||
| 84 | " key = (key, int(censorship))\n", |
||
| 85 | " slide_data.at[i, 'label'] = label_dict[key]\n", |
||
| 86 | "\n", |
||
| 87 | "bins = q_bins\n", |
||
| 88 | "num_classes=len(label_dict)\n", |
||
| 89 | "patients_df = slide_data.drop_duplicates(['case_id'])\n", |
||
| 90 | "patient_data = {'case_id':patients_df['case_id'].values, 'label':patients_df['label'].values}\n", |
||
| 91 | "\n", |
||
| 92 | "new_cols = list(slide_data.columns[-2:]) + list(slide_data.columns[:-2])\n", |
||
| 93 | "slide_data = slide_data[new_cols]\n", |
||
| 94 | "metadata = slide_data.columns[:11]" |
||
| 95 | ] |
||
| 96 | }, |
||
| 97 | { |
||
| 98 | "cell_type": "code", |
||
| 99 | "execution_count": 5, |
||
| 100 | "metadata": {}, |
||
| 101 | "outputs": [], |
||
| 102 | "source": [ |
||
| 103 | "from sklearn.pipeline import Pipeline\n", |
||
| 104 | "from sklearn.decomposition import PCA\n", |
||
| 105 | "from sklearn.preprocessing import StandardScaler\n", |
||
| 106 | "\n", |
||
| 107 | "\n", |
||
| 108 | "def series_intersection(s1, s2):\n", |
||
| 109 | " return pd.Series(list(set(s1) & set(s2)))\n", |
||
| 110 | "\n", |
||
| 111 | "genomic_features = slide_data.drop(metadata, axis=1)\n", |
||
| 112 | "scaler_omic = StandardScaler().fit(genomic_features)" |
||
| 113 | ] |
||
| 114 | }, |
||
| 115 | { |
||
| 116 | "cell_type": "code", |
||
| 117 | "execution_count": 9, |
||
| 118 | "metadata": {}, |
||
| 119 | "outputs": [ |
||
| 120 | { |
||
| 121 | "name": "stderr", |
||
| 122 | "output_type": "stream", |
||
| 123 | "text": [ |
||
| 124 | "/home/mahmoodlab/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3071: DtypeWarning: Columns (2) have mixed types.Specify dtype option on import or set low_memory=False.\n", |
||
| 125 | " has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n" |
||
| 126 | ] |
||
| 127 | } |
||
| 128 | ], |
||
| 129 | "source": [ |
||
| 130 | "import os\n", |
||
| 131 | "from os.path import join\n", |
||
| 132 | "\n", |
||
| 133 | "import pandas as pd\n", |
||
| 134 | "import numpy as np\n", |
||
| 135 | "\n", |
||
| 136 | "signatures = pd.read_csv('./signatures.csv')\n", |
||
| 137 | "slide_df = pd.read_csv('./tcga_gbmlgg_all_clean.csv.zip')" |
||
| 138 | ] |
||
| 139 | }, |
||
| 140 | { |
||
| 141 | "cell_type": "code", |
||
| 142 | "execution_count": 43, |
||
| 143 | "metadata": {}, |
||
| 144 | "outputs": [ |
||
| 145 | { |
||
| 146 | "data": { |
||
| 147 | "text/html": [ |
||
| 148 | "<div>\n", |
||
| 149 | "<style scoped>\n", |
||
| 150 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 151 | " vertical-align: middle;\n", |
||
| 152 | " }\n", |
||
| 153 | "\n", |
||
| 154 | " .dataframe tbody tr th {\n", |
||
| 155 | " vertical-align: top;\n", |
||
| 156 | " }\n", |
||
| 157 | "\n", |
||
| 158 | " .dataframe thead th {\n", |
||
| 159 | " text-align: right;\n", |
||
| 160 | " }\n", |
||
| 161 | "</style>\n", |
||
| 162 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 163 | " <thead>\n", |
||
| 164 | " <tr style=\"text-align: right;\">\n", |
||
| 165 | " <th></th>\n", |
||
| 166 | " <th>Unnamed: 0</th>\n", |
||
| 167 | " <th>Unnamed: 0.1</th>\n", |
||
| 168 | " <th>case_id</th>\n", |
||
| 169 | " <th>slide_id</th>\n", |
||
| 170 | " <th>site</th>\n", |
||
| 171 | " <th>is_female</th>\n", |
||
| 172 | " <th>oncotree_code</th>\n", |
||
| 173 | " <th>age</th>\n", |
||
| 174 | " <th>survival_months</th>\n", |
||
| 175 | " <th>censorship</th>\n", |
||
| 176 | " <th>...</th>\n", |
||
| 177 | " <th>ZSCAN10_rnaseq</th>\n", |
||
| 178 | " <th>ZSCAN12_rnaseq</th>\n", |
||
| 179 | " <th>ZSCAN20_rnaseq</th>\n", |
||
| 180 | " <th>ZSCAN21_rnaseq</th>\n", |
||
| 181 | " <th>ZSCAN22_rnaseq</th>\n", |
||
| 182 | " <th>ZSCAN2_rnaseq</th>\n", |
||
| 183 | " <th>ZSCAN9_rnaseq</th>\n", |
||
| 184 | " <th>ZXDA_rnaseq</th>\n", |
||
| 185 | " <th>ZXDB_rnaseq</th>\n", |
||
| 186 | " <th>ZXDC_rnaseq</th>\n", |
||
| 187 | " </tr>\n", |
||
| 188 | " </thead>\n", |
||
| 189 | " <tbody>\n", |
||
| 190 | " <tr>\n", |
||
| 191 | " <th>0</th>\n", |
||
| 192 | " <td>0</td>\n", |
||
| 193 | " <td>0</td>\n", |
||
| 194 | " <td>TCGA-02-0047</td>\n", |
||
| 195 | " <td>TCGA-02-0047-01Z-00-DX1.4755D138-5842-4159-848...</td>\n", |
||
| 196 | " <td>2</td>\n", |
||
| 197 | " <td>0.0</td>\n", |
||
| 198 | " <td>GBM</td>\n", |
||
| 199 | " <td>78.0</td>\n", |
||
| 200 | " <td>14.72</td>\n", |
||
| 201 | " <td>0.0</td>\n", |
||
| 202 | " <td>...</td>\n", |
||
| 203 | " <td>-0.1599</td>\n", |
||
| 204 | " <td>-0.59540</td>\n", |
||
| 205 | " <td>0.0813</td>\n", |
||
| 206 | " <td>-1.16960</td>\n", |
||
| 207 | " <td>-0.1728</td>\n", |
||
| 208 | " <td>-0.1144</td>\n", |
||
| 209 | " <td>-0.4155</td>\n", |
||
| 210 | " <td>0.4046</td>\n", |
||
| 211 | " <td>-0.01680</td>\n", |
||
| 212 | " <td>0.3026</td>\n", |
||
| 213 | " </tr>\n", |
||
| 214 | " <tr>\n", |
||
| 215 | " <th>1</th>\n", |
||
| 216 | " <td>1</td>\n", |
||
| 217 | " <td>1</td>\n", |
||
| 218 | " <td>TCGA-06-0125</td>\n", |
||
| 219 | " <td>TCGA-06-0125-01Z-00-DX1.8e0915b2-8dc3-4753-806...</td>\n", |
||
| 220 | " <td>6</td>\n", |
||
| 221 | " <td>1.0</td>\n", |
||
| 222 | " <td>GBM</td>\n", |
||
| 223 | " <td>63.0</td>\n", |
||
| 224 | " <td>47.57</td>\n", |
||
| 225 | " <td>0.0</td>\n", |
||
| 226 | " <td>...</td>\n", |
||
| 227 | " <td>0.4608</td>\n", |
||
| 228 | " <td>0.52815</td>\n", |
||
| 229 | " <td>1.2580</td>\n", |
||
| 230 | " <td>1.41685</td>\n", |
||
| 231 | " <td>2.4839</td>\n", |
||
| 232 | " <td>-0.2388</td>\n", |
||
| 233 | " <td>0.9025</td>\n", |
||
| 234 | " <td>0.3242</td>\n", |
||
| 235 | " <td>1.01905</td>\n", |
||
| 236 | " <td>0.2265</td>\n", |
||
| 237 | " </tr>\n", |
||
| 238 | " <tr>\n", |
||
| 239 | " <th>2</th>\n", |
||
| 240 | " <td>2</td>\n", |
||
| 241 | " <td>2</td>\n", |
||
| 242 | " <td>TCGA-06-0125</td>\n", |
||
| 243 | " <td>TCGA-06-0125-01Z-00-DX2.4f9cef92-2bdb-480d-870...</td>\n", |
||
| 244 | " <td>6</td>\n", |
||
| 245 | " <td>1.0</td>\n", |
||
| 246 | " <td>GBM</td>\n", |
||
| 247 | " <td>63.0</td>\n", |
||
| 248 | " <td>47.57</td>\n", |
||
| 249 | " <td>0.0</td>\n", |
||
| 250 | " <td>...</td>\n", |
||
| 251 | " <td>0.4608</td>\n", |
||
| 252 | " <td>0.52815</td>\n", |
||
| 253 | " <td>1.2580</td>\n", |
||
| 254 | " <td>1.41685</td>\n", |
||
| 255 | " <td>2.4839</td>\n", |
||
| 256 | " <td>-0.2388</td>\n", |
||
| 257 | " <td>0.9025</td>\n", |
||
| 258 | " <td>0.3242</td>\n", |
||
| 259 | " <td>1.01905</td>\n", |
||
| 260 | " <td>0.2265</td>\n", |
||
| 261 | " </tr>\n", |
||
| 262 | " <tr>\n", |
||
| 263 | " <th>3</th>\n", |
||
| 264 | " <td>3</td>\n", |
||
| 265 | " <td>3</td>\n", |
||
| 266 | " <td>TCGA-06-0129</td>\n", |
||
| 267 | " <td>TCGA-06-0129-01Z-00-DX1.b7bddf7d-f39e-45e7-a78...</td>\n", |
||
| 268 | " <td>6</td>\n", |
||
| 269 | " <td>0.0</td>\n", |
||
| 270 | " <td>GBM</td>\n", |
||
| 271 | " <td>30.0</td>\n", |
||
| 272 | " <td>33.64</td>\n", |
||
| 273 | " <td>0.0</td>\n", |
||
| 274 | " <td>...</td>\n", |
||
| 275 | " <td>-0.2960</td>\n", |
||
| 276 | " <td>-0.75980</td>\n", |
||
| 277 | " <td>1.2706</td>\n", |
||
| 278 | " <td>-0.14840</td>\n", |
||
| 279 | " <td>1.4803</td>\n", |
||
| 280 | " <td>1.5796</td>\n", |
||
| 281 | " <td>1.0245</td>\n", |
||
| 282 | " <td>1.0492</td>\n", |
||
| 283 | " <td>5.78560</td>\n", |
||
| 284 | " <td>1.7766</td>\n", |
||
| 285 | " </tr>\n", |
||
| 286 | " <tr>\n", |
||
| 287 | " <th>4</th>\n", |
||
| 288 | " <td>4</td>\n", |
||
| 289 | " <td>4</td>\n", |
||
| 290 | " <td>TCGA-06-0129</td>\n", |
||
| 291 | " <td>TCGA-06-0129-01Z-00-DX2.1ea78b46-1dc7-44d8-81b...</td>\n", |
||
| 292 | " <td>6</td>\n", |
||
| 293 | " <td>0.0</td>\n", |
||
| 294 | " <td>GBM</td>\n", |
||
| 295 | " <td>30.0</td>\n", |
||
| 296 | " <td>33.64</td>\n", |
||
| 297 | " <td>0.0</td>\n", |
||
| 298 | " <td>...</td>\n", |
||
| 299 | " <td>-0.2960</td>\n", |
||
| 300 | " <td>-0.75980</td>\n", |
||
| 301 | " <td>1.2706</td>\n", |
||
| 302 | " <td>-0.14840</td>\n", |
||
| 303 | " <td>1.4803</td>\n", |
||
| 304 | " <td>1.5796</td>\n", |
||
| 305 | " <td>1.0245</td>\n", |
||
| 306 | " <td>1.0492</td>\n", |
||
| 307 | " <td>5.78560</td>\n", |
||
| 308 | " <td>1.7766</td>\n", |
||
| 309 | " </tr>\n", |
||
| 310 | " <tr>\n", |
||
| 311 | " <th>...</th>\n", |
||
| 312 | " <td>...</td>\n", |
||
| 313 | " <td>...</td>\n", |
||
| 314 | " <td>...</td>\n", |
||
| 315 | " <td>...</td>\n", |
||
| 316 | " <td>...</td>\n", |
||
| 317 | " <td>...</td>\n", |
||
| 318 | " <td>...</td>\n", |
||
| 319 | " <td>...</td>\n", |
||
| 320 | " <td>...</td>\n", |
||
| 321 | " <td>...</td>\n", |
||
| 322 | " <td>...</td>\n", |
||
| 323 | " <td>...</td>\n", |
||
| 324 | " <td>...</td>\n", |
||
| 325 | " <td>...</td>\n", |
||
| 326 | " <td>...</td>\n", |
||
| 327 | " <td>...</td>\n", |
||
| 328 | " <td>...</td>\n", |
||
| 329 | " <td>...</td>\n", |
||
| 330 | " <td>...</td>\n", |
||
| 331 | " <td>...</td>\n", |
||
| 332 | " <td>...</td>\n", |
||
| 333 | " </tr>\n", |
||
| 334 | " <tr>\n", |
||
| 335 | " <th>1037</th>\n", |
||
| 336 | " <td>1037</td>\n", |
||
| 337 | " <td>1037</td>\n", |
||
| 338 | " <td>TCGA-WY-A85A</td>\n", |
||
| 339 | " <td>TCGA-WY-A85A-01Z-00-DX1.CB302B89-F89A-40FD-A7D...</td>\n", |
||
| 340 | " <td>WY</td>\n", |
||
| 341 | " <td>0.0</td>\n", |
||
| 342 | " <td>ASTR</td>\n", |
||
| 343 | " <td>20.0</td>\n", |
||
| 344 | " <td>43.36</td>\n", |
||
| 345 | " <td>1.0</td>\n", |
||
| 346 | " <td>...</td>\n", |
||
| 347 | " <td>-0.2997</td>\n", |
||
| 348 | " <td>-0.67560</td>\n", |
||
| 349 | " <td>0.2714</td>\n", |
||
| 350 | " <td>0.36210</td>\n", |
||
| 351 | " <td>-0.2401</td>\n", |
||
| 352 | " <td>1.4333</td>\n", |
||
| 353 | " <td>0.2715</td>\n", |
||
| 354 | " <td>-0.5415</td>\n", |
||
| 355 | " <td>-0.69620</td>\n", |
||
| 356 | " <td>-0.1123</td>\n", |
||
| 357 | " </tr>\n", |
||
| 358 | " <tr>\n", |
||
| 359 | " <th>1038</th>\n", |
||
| 360 | " <td>1038</td>\n", |
||
| 361 | " <td>1038</td>\n", |
||
| 362 | " <td>TCGA-WY-A85B</td>\n", |
||
| 363 | " <td>TCGA-WY-A85B-01Z-00-DX1.1E4B796A-A1E3-45F9-807...</td>\n", |
||
| 364 | " <td>WY</td>\n", |
||
| 365 | " <td>0.0</td>\n", |
||
| 366 | " <td>ASTR</td>\n", |
||
| 367 | " <td>24.0</td>\n", |
||
| 368 | " <td>45.76</td>\n", |
||
| 369 | " <td>1.0</td>\n", |
||
| 370 | " <td>...</td>\n", |
||
| 371 | " <td>-0.0678</td>\n", |
||
| 372 | " <td>0.30360</td>\n", |
||
| 373 | " <td>0.3361</td>\n", |
||
| 374 | " <td>1.21610</td>\n", |
||
| 375 | " <td>0.9365</td>\n", |
||
| 376 | " <td>1.4954</td>\n", |
||
| 377 | " <td>1.4201</td>\n", |
||
| 378 | " <td>-0.3525</td>\n", |
||
| 379 | " <td>0.52860</td>\n", |
||
| 380 | " <td>0.1971</td>\n", |
||
| 381 | " </tr>\n", |
||
| 382 | " <tr>\n", |
||
| 383 | " <th>1039</th>\n", |
||
| 384 | " <td>1039</td>\n", |
||
| 385 | " <td>1039</td>\n", |
||
| 386 | " <td>TCGA-WY-A85C</td>\n", |
||
| 387 | " <td>TCGA-WY-A85C-01Z-00-DX1.E0A6429A-91B3-4FFE-9FF...</td>\n", |
||
| 388 | " <td>WY</td>\n", |
||
| 389 | " <td>0.0</td>\n", |
||
| 390 | " <td>ASTR</td>\n", |
||
| 391 | " <td>36.0</td>\n", |
||
| 392 | " <td>46.85</td>\n", |
||
| 393 | " <td>1.0</td>\n", |
||
| 394 | " <td>...</td>\n", |
||
| 395 | " <td>0.0320</td>\n", |
||
| 396 | " <td>-1.01940</td>\n", |
||
| 397 | " <td>0.6582</td>\n", |
||
| 398 | " <td>2.55740</td>\n", |
||
| 399 | " <td>1.4708</td>\n", |
||
| 400 | " <td>0.8381</td>\n", |
||
| 401 | " <td>2.9481</td>\n", |
||
| 402 | " <td>0.1252</td>\n", |
||
| 403 | " <td>0.75300</td>\n", |
||
| 404 | " <td>0.9603</td>\n", |
||
| 405 | " </tr>\n", |
||
| 406 | " <tr>\n", |
||
| 407 | " <th>1040</th>\n", |
||
| 408 | " <td>1040</td>\n", |
||
| 409 | " <td>1040</td>\n", |
||
| 410 | " <td>TCGA-WY-A85D</td>\n", |
||
| 411 | " <td>TCGA-WY-A85D-01Z-00-DX1.FB8C252B-7A88-4B14-B3C...</td>\n", |
||
| 412 | " <td>WY</td>\n", |
||
| 413 | " <td>0.0</td>\n", |
||
| 414 | " <td>OAST</td>\n", |
||
| 415 | " <td>60.0</td>\n", |
||
| 416 | " <td>37.68</td>\n", |
||
| 417 | " <td>1.0</td>\n", |
||
| 418 | " <td>...</td>\n", |
||
| 419 | " <td>-0.3021</td>\n", |
||
| 420 | " <td>-0.34820</td>\n", |
||
| 421 | " <td>-0.4824</td>\n", |
||
| 422 | " <td>1.57910</td>\n", |
||
| 423 | " <td>0.0187</td>\n", |
||
| 424 | " <td>-0.7983</td>\n", |
||
| 425 | " <td>1.4101</td>\n", |
||
| 426 | " <td>-1.0976</td>\n", |
||
| 427 | " <td>-1.00950</td>\n", |
||
| 428 | " <td>0.5940</td>\n", |
||
| 429 | " </tr>\n", |
||
| 430 | " <tr>\n", |
||
| 431 | " <th>1041</th>\n", |
||
| 432 | " <td>1041</td>\n", |
||
| 433 | " <td>1041</td>\n", |
||
| 434 | " <td>TCGA-WY-A85E</td>\n", |
||
| 435 | " <td>TCGA-WY-A85E-01Z-00-DX1.AA7A4C1F-99AA-490D-B6D...</td>\n", |
||
| 436 | " <td>WY</td>\n", |
||
| 437 | " <td>1.0</td>\n", |
||
| 438 | " <td>OAST</td>\n", |
||
| 439 | " <td>48.0</td>\n", |
||
| 440 | " <td>20.80</td>\n", |
||
| 441 | " <td>1.0</td>\n", |
||
| 442 | " <td>...</td>\n", |
||
| 443 | " <td>-0.2576</td>\n", |
||
| 444 | " <td>0.89960</td>\n", |
||
| 445 | " <td>-0.7533</td>\n", |
||
| 446 | " <td>1.42710</td>\n", |
||
| 447 | " <td>-0.6667</td>\n", |
||
| 448 | " <td>0.8354</td>\n", |
||
| 449 | " <td>1.2988</td>\n", |
||
| 450 | " <td>-0.4902</td>\n", |
||
| 451 | " <td>-0.42940</td>\n", |
||
| 452 | " <td>-2.0717</td>\n", |
||
| 453 | " </tr>\n", |
||
| 454 | " </tbody>\n", |
||
| 455 | "</table>\n", |
||
| 456 | "<p>1042 rows × 2842 columns</p>\n", |
||
| 457 | "</div>" |
||
| 458 | ], |
||
| 459 | "text/plain": [ |
||
| 460 | " Unnamed: 0 Unnamed: 0.1 case_id \\\n", |
||
| 461 | "0 0 0 TCGA-02-0047 \n", |
||
| 462 | "1 1 1 TCGA-06-0125 \n", |
||
| 463 | "2 2 2 TCGA-06-0125 \n", |
||
| 464 | "3 3 3 TCGA-06-0129 \n", |
||
| 465 | "4 4 4 TCGA-06-0129 \n", |
||
| 466 | "... ... ... ... \n", |
||
| 467 | "1037 1037 1037 TCGA-WY-A85A \n", |
||
| 468 | "1038 1038 1038 TCGA-WY-A85B \n", |
||
| 469 | "1039 1039 1039 TCGA-WY-A85C \n", |
||
| 470 | "1040 1040 1040 TCGA-WY-A85D \n", |
||
| 471 | "1041 1041 1041 TCGA-WY-A85E \n", |
||
| 472 | "\n", |
||
| 473 | " slide_id site is_female \\\n", |
||
| 474 | "0 TCGA-02-0047-01Z-00-DX1.4755D138-5842-4159-848... 2 0.0 \n", |
||
| 475 | "1 TCGA-06-0125-01Z-00-DX1.8e0915b2-8dc3-4753-806... 6 1.0 \n", |
||
| 476 | "2 TCGA-06-0125-01Z-00-DX2.4f9cef92-2bdb-480d-870... 6 1.0 \n", |
||
| 477 | "3 TCGA-06-0129-01Z-00-DX1.b7bddf7d-f39e-45e7-a78... 6 0.0 \n", |
||
| 478 | "4 TCGA-06-0129-01Z-00-DX2.1ea78b46-1dc7-44d8-81b... 6 0.0 \n", |
||
| 479 | "... ... ... ... \n", |
||
| 480 | "1037 TCGA-WY-A85A-01Z-00-DX1.CB302B89-F89A-40FD-A7D... WY 0.0 \n", |
||
| 481 | "1038 TCGA-WY-A85B-01Z-00-DX1.1E4B796A-A1E3-45F9-807... WY 0.0 \n", |
||
| 482 | "1039 TCGA-WY-A85C-01Z-00-DX1.E0A6429A-91B3-4FFE-9FF... WY 0.0 \n", |
||
| 483 | "1040 TCGA-WY-A85D-01Z-00-DX1.FB8C252B-7A88-4B14-B3C... WY 0.0 \n", |
||
| 484 | "1041 TCGA-WY-A85E-01Z-00-DX1.AA7A4C1F-99AA-490D-B6D... WY 1.0 \n", |
||
| 485 | "\n", |
||
| 486 | " oncotree_code age survival_months censorship ... ZSCAN10_rnaseq \\\n", |
||
| 487 | "0 GBM 78.0 14.72 0.0 ... -0.1599 \n", |
||
| 488 | "1 GBM 63.0 47.57 0.0 ... 0.4608 \n", |
||
| 489 | "2 GBM 63.0 47.57 0.0 ... 0.4608 \n", |
||
| 490 | "3 GBM 30.0 33.64 0.0 ... -0.2960 \n", |
||
| 491 | "4 GBM 30.0 33.64 0.0 ... -0.2960 \n", |
||
| 492 | "... ... ... ... ... ... ... \n", |
||
| 493 | "1037 ASTR 20.0 43.36 1.0 ... -0.2997 \n", |
||
| 494 | "1038 ASTR 24.0 45.76 1.0 ... -0.0678 \n", |
||
| 495 | "1039 ASTR 36.0 46.85 1.0 ... 0.0320 \n", |
||
| 496 | "1040 OAST 60.0 37.68 1.0 ... -0.3021 \n", |
||
| 497 | "1041 OAST 48.0 20.80 1.0 ... -0.2576 \n", |
||
| 498 | "\n", |
||
| 499 | " ZSCAN12_rnaseq ZSCAN20_rnaseq ZSCAN21_rnaseq ZSCAN22_rnaseq \\\n", |
||
| 500 | "0 -0.59540 0.0813 -1.16960 -0.1728 \n", |
||
| 501 | "1 0.52815 1.2580 1.41685 2.4839 \n", |
||
| 502 | "2 0.52815 1.2580 1.41685 2.4839 \n", |
||
| 503 | "3 -0.75980 1.2706 -0.14840 1.4803 \n", |
||
| 504 | "4 -0.75980 1.2706 -0.14840 1.4803 \n", |
||
| 505 | "... ... ... ... ... \n", |
||
| 506 | "1037 -0.67560 0.2714 0.36210 -0.2401 \n", |
||
| 507 | "1038 0.30360 0.3361 1.21610 0.9365 \n", |
||
| 508 | "1039 -1.01940 0.6582 2.55740 1.4708 \n", |
||
| 509 | "1040 -0.34820 -0.4824 1.57910 0.0187 \n", |
||
| 510 | "1041 0.89960 -0.7533 1.42710 -0.6667 \n", |
||
| 511 | "\n", |
||
| 512 | " ZSCAN2_rnaseq ZSCAN9_rnaseq ZXDA_rnaseq ZXDB_rnaseq ZXDC_rnaseq \n", |
||
| 513 | "0 -0.1144 -0.4155 0.4046 -0.01680 0.3026 \n", |
||
| 514 | "1 -0.2388 0.9025 0.3242 1.01905 0.2265 \n", |
||
| 515 | "2 -0.2388 0.9025 0.3242 1.01905 0.2265 \n", |
||
| 516 | "3 1.5796 1.0245 1.0492 5.78560 1.7766 \n", |
||
| 517 | "4 1.5796 1.0245 1.0492 5.78560 1.7766 \n", |
||
| 518 | "... ... ... ... ... ... \n", |
||
| 519 | "1037 1.4333 0.2715 -0.5415 -0.69620 -0.1123 \n", |
||
| 520 | "1038 1.4954 1.4201 -0.3525 0.52860 0.1971 \n", |
||
| 521 | "1039 0.8381 2.9481 0.1252 0.75300 0.9603 \n", |
||
| 522 | "1040 -0.7983 1.4101 -1.0976 -1.00950 0.5940 \n", |
||
| 523 | "1041 0.8354 1.2988 -0.4902 -0.42940 -2.0717 \n", |
||
| 524 | "\n", |
||
| 525 | "[1042 rows x 2842 columns]" |
||
| 526 | ] |
||
| 527 | }, |
||
| 528 | "execution_count": 43, |
||
| 529 | "metadata": {}, |
||
| 530 | "output_type": "execute_result" |
||
| 531 | } |
||
| 532 | ], |
||
| 533 | "source": [ |
||
| 534 | "pd.read_csv(fname)" |
||
| 535 | ] |
||
| 536 | }, |
||
| 537 | { |
||
| 538 | "cell_type": "code", |
||
| 539 | "execution_count": 76, |
||
| 540 | "metadata": {}, |
||
| 541 | "outputs": [], |
||
| 542 | "source": [ |
||
| 543 | "omic_from_signatures = []\n", |
||
| 544 | "for col in signatures.columns:\n", |
||
| 545 | " omic = signatures[col].dropna().unique()\n", |
||
| 546 | " omic_from_signatures.append(omic)\n", |
||
| 547 | "\n", |
||
| 548 | "omic_from_signatures = np.concatenate(omic_from_signatures)" |
||
| 549 | ] |
||
| 550 | }, |
||
| 551 | { |
||
| 552 | "cell_type": "code", |
||
| 553 | "execution_count": 44, |
||
| 554 | "metadata": {}, |
||
| 555 | "outputs": [ |
||
| 556 | { |
||
| 557 | "name": "stderr", |
||
| 558 | "output_type": "stream", |
||
| 559 | "text": [ |
||
| 560 | "/home/mahmoodlab/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3071: DtypeWarning: Columns (3) have mixed types.Specify dtype option on import or set low_memory=False.\n", |
||
| 561 | " has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n" |
||
| 562 | ] |
||
| 563 | } |
||
| 564 | ], |
||
| 565 | "source": [ |
||
| 566 | "for fname in os.listdir('./'):\n", |
||
| 567 | " if fname.endswith('.csv.zip'):\n", |
||
| 568 | " slide_df = pd.read_csv(fname)" |
||
| 569 | ] |
||
| 570 | }, |
||
| 571 | { |
||
| 572 | "cell_type": "code", |
||
| 573 | "execution_count": 81, |
||
| 574 | "metadata": {}, |
||
| 575 | "outputs": [ |
||
| 576 | { |
||
| 577 | "data": { |
||
| 578 | "text/html": [ |
||
| 579 | "<div>\n", |
||
| 580 | "<style scoped>\n", |
||
| 581 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 582 | " vertical-align: middle;\n", |
||
| 583 | " }\n", |
||
| 584 | "\n", |
||
| 585 | " .dataframe tbody tr th {\n", |
||
| 586 | " vertical-align: top;\n", |
||
| 587 | " }\n", |
||
| 588 | "\n", |
||
| 589 | " .dataframe thead th {\n", |
||
| 590 | " text-align: right;\n", |
||
| 591 | " }\n", |
||
| 592 | "</style>\n", |
||
| 593 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 594 | " <thead>\n", |
||
| 595 | " <tr style=\"text-align: right;\">\n", |
||
| 596 | " <th></th>\n", |
||
| 597 | " <th>Unnamed: 0</th>\n", |
||
| 598 | " <th>Unnamed: 0.1</th>\n", |
||
| 599 | " <th>case_id</th>\n", |
||
| 600 | " <th>slide_id</th>\n", |
||
| 601 | " <th>site</th>\n", |
||
| 602 | " <th>is_female</th>\n", |
||
| 603 | " <th>oncotree_code</th>\n", |
||
| 604 | " <th>age</th>\n", |
||
| 605 | " <th>survival_months</th>\n", |
||
| 606 | " <th>censorship</th>\n", |
||
| 607 | " <th>...</th>\n", |
||
| 608 | " <th>ZSCAN10_rnaseq</th>\n", |
||
| 609 | " <th>ZSCAN12_rnaseq</th>\n", |
||
| 610 | " <th>ZSCAN20_rnaseq</th>\n", |
||
| 611 | " <th>ZSCAN21_rnaseq</th>\n", |
||
| 612 | " <th>ZSCAN22_rnaseq</th>\n", |
||
| 613 | " <th>ZSCAN2_rnaseq</th>\n", |
||
| 614 | " <th>ZSCAN9_rnaseq</th>\n", |
||
| 615 | " <th>ZXDA_rnaseq</th>\n", |
||
| 616 | " <th>ZXDB_rnaseq</th>\n", |
||
| 617 | " <th>ZXDC_rnaseq</th>\n", |
||
| 618 | " </tr>\n", |
||
| 619 | " </thead>\n", |
||
| 620 | " <tbody>\n", |
||
| 621 | " <tr>\n", |
||
| 622 | " <th>0</th>\n", |
||
| 623 | " <td>0</td>\n", |
||
| 624 | " <td>0</td>\n", |
||
| 625 | " <td>TCGA-05-4249</td>\n", |
||
| 626 | " <td>TCGA-05-4249-01Z-00-DX1.9fce0297-cc19-4c04-872...</td>\n", |
||
| 627 | " <td>5</td>\n", |
||
| 628 | " <td>0.0</td>\n", |
||
| 629 | " <td>LUAD</td>\n", |
||
| 630 | " <td>67.0</td>\n", |
||
| 631 | " <td>50.03</td>\n", |
||
| 632 | " <td>1.0</td>\n", |
||
| 633 | " <td>...</td>\n", |
||
| 634 | " <td>-0.1238</td>\n", |
||
| 635 | " <td>0.7530</td>\n", |
||
| 636 | " <td>0.6552</td>\n", |
||
| 637 | " <td>-1.0013</td>\n", |
||
| 638 | " <td>0.2353</td>\n", |
||
| 639 | " <td>2.6532</td>\n", |
||
| 640 | " <td>1.1103</td>\n", |
||
| 641 | " <td>0.6149</td>\n", |
||
| 642 | " <td>0.5725</td>\n", |
||
| 643 | " <td>0.2889</td>\n", |
||
| 644 | " </tr>\n", |
||
| 645 | " <tr>\n", |
||
| 646 | " <th>1</th>\n", |
||
| 647 | " <td>1</td>\n", |
||
| 648 | " <td>1</td>\n", |
||
| 649 | " <td>TCGA-05-4250</td>\n", |
||
| 650 | " <td>TCGA-05-4250-01Z-00-DX1.90f67fdf-dff9-46ca-af7...</td>\n", |
||
| 651 | " <td>5</td>\n", |
||
| 652 | " <td>1.0</td>\n", |
||
| 653 | " <td>LUAD</td>\n", |
||
| 654 | " <td>79.0</td>\n", |
||
| 655 | " <td>3.98</td>\n", |
||
| 656 | " <td>0.0</td>\n", |
||
| 657 | " <td>...</td>\n", |
||
| 658 | " <td>-0.1238</td>\n", |
||
| 659 | " <td>0.4810</td>\n", |
||
| 660 | " <td>-0.8255</td>\n", |
||
| 661 | " <td>0.2825</td>\n", |
||
| 662 | " <td>-1.2502</td>\n", |
||
| 663 | " <td>-0.9024</td>\n", |
||
| 664 | " <td>-0.1472</td>\n", |
||
| 665 | " <td>0.5118</td>\n", |
||
| 666 | " <td>-0.1673</td>\n", |
||
| 667 | " <td>-0.8006</td>\n", |
||
| 668 | " </tr>\n", |
||
| 669 | " <tr>\n", |
||
| 670 | " <th>2</th>\n", |
||
| 671 | " <td>2</td>\n", |
||
| 672 | " <td>2</td>\n", |
||
| 673 | " <td>TCGA-05-4382</td>\n", |
||
| 674 | " <td>TCGA-05-4382-01Z-00-DX1.76b49a4c-dbbb-48b0-b67...</td>\n", |
||
| 675 | " <td>5</td>\n", |
||
| 676 | " <td>0.0</td>\n", |
||
| 677 | " <td>LUAD</td>\n", |
||
| 678 | " <td>68.0</td>\n", |
||
| 679 | " <td>19.94</td>\n", |
||
| 680 | " <td>1.0</td>\n", |
||
| 681 | " <td>...</td>\n", |
||
| 682 | " <td>0.3265</td>\n", |
||
| 683 | " <td>0.4462</td>\n", |
||
| 684 | " <td>1.1847</td>\n", |
||
| 685 | " <td>0.8765</td>\n", |
||
| 686 | " <td>-0.7999</td>\n", |
||
| 687 | " <td>1.7566</td>\n", |
||
| 688 | " <td>1.1757</td>\n", |
||
| 689 | " <td>-0.4399</td>\n", |
||
| 690 | " <td>-0.2751</td>\n", |
||
| 691 | " <td>-0.4668</td>\n", |
||
| 692 | " </tr>\n", |
||
| 693 | " <tr>\n", |
||
| 694 | " <th>3</th>\n", |
||
| 695 | " <td>3</td>\n", |
||
| 696 | " <td>3</td>\n", |
||
| 697 | " <td>TCGA-05-4384</td>\n", |
||
| 698 | " <td>TCGA-05-4384-01Z-00-DX1.CA68BF29-BBE3-4C8E-B48...</td>\n", |
||
| 699 | " <td>5</td>\n", |
||
| 700 | " <td>0.0</td>\n", |
||
| 701 | " <td>LUAD</td>\n", |
||
| 702 | " <td>66.0</td>\n", |
||
| 703 | " <td>13.99</td>\n", |
||
| 704 | " <td>1.0</td>\n", |
||
| 705 | " <td>...</td>\n", |
||
| 706 | " <td>-0.1238</td>\n", |
||
| 707 | " <td>-0.0369</td>\n", |
||
| 708 | " <td>0.5766</td>\n", |
||
| 709 | " <td>0.0083</td>\n", |
||
| 710 | " <td>0.1344</td>\n", |
||
| 711 | " <td>0.8299</td>\n", |
||
| 712 | " <td>0.6599</td>\n", |
||
| 713 | " <td>1.4844</td>\n", |
||
| 714 | " <td>0.9748</td>\n", |
||
| 715 | " <td>0.7481</td>\n", |
||
| 716 | " </tr>\n", |
||
| 717 | " <tr>\n", |
||
| 718 | " <th>4</th>\n", |
||
| 719 | " <td>4</td>\n", |
||
| 720 | " <td>4</td>\n", |
||
| 721 | " <td>TCGA-05-4390</td>\n", |
||
| 722 | " <td>TCGA-05-4390-01Z-00-DX1.858E64DF-DD3E-4F43-B7C...</td>\n", |
||
| 723 | " <td>5</td>\n", |
||
| 724 | " <td>1.0</td>\n", |
||
| 725 | " <td>LUAD</td>\n", |
||
| 726 | " <td>58.0</td>\n", |
||
| 727 | " <td>36.99</td>\n", |
||
| 728 | " <td>1.0</td>\n", |
||
| 729 | " <td>...</td>\n", |
||
| 730 | " <td>-0.1238</td>\n", |
||
| 731 | " <td>0.4751</td>\n", |
||
| 732 | " <td>1.2404</td>\n", |
||
| 733 | " <td>0.6932</td>\n", |
||
| 734 | " <td>-0.2792</td>\n", |
||
| 735 | " <td>2.1326</td>\n", |
||
| 736 | " <td>0.1621</td>\n", |
||
| 737 | " <td>-0.0462</td>\n", |
||
| 738 | " <td>1.8418</td>\n", |
||
| 739 | " <td>-0.9922</td>\n", |
||
| 740 | " </tr>\n", |
||
| 741 | " <tr>\n", |
||
| 742 | " <th>...</th>\n", |
||
| 743 | " <td>...</td>\n", |
||
| 744 | " <td>...</td>\n", |
||
| 745 | " <td>...</td>\n", |
||
| 746 | " <td>...</td>\n", |
||
| 747 | " <td>...</td>\n", |
||
| 748 | " <td>...</td>\n", |
||
| 749 | " <td>...</td>\n", |
||
| 750 | " <td>...</td>\n", |
||
| 751 | " <td>...</td>\n", |
||
| 752 | " <td>...</td>\n", |
||
| 753 | " <td>...</td>\n", |
||
| 754 | " <td>...</td>\n", |
||
| 755 | " <td>...</td>\n", |
||
| 756 | " <td>...</td>\n", |
||
| 757 | " <td>...</td>\n", |
||
| 758 | " <td>...</td>\n", |
||
| 759 | " <td>...</td>\n", |
||
| 760 | " <td>...</td>\n", |
||
| 761 | " <td>...</td>\n", |
||
| 762 | " <td>...</td>\n", |
||
| 763 | " <td>...</td>\n", |
||
| 764 | " </tr>\n", |
||
| 765 | " <tr>\n", |
||
| 766 | " <th>511</th>\n", |
||
| 767 | " <td>511</td>\n", |
||
| 768 | " <td>511</td>\n", |
||
| 769 | " <td>TCGA-NJ-A55O</td>\n", |
||
| 770 | " <td>TCGA-NJ-A55O-01Z-00-DX1.8E23C821-B8BB-4D89-9E3...</td>\n", |
||
| 771 | " <td>NJ</td>\n", |
||
| 772 | " <td>1.0</td>\n", |
||
| 773 | " <td>LUAD</td>\n", |
||
| 774 | " <td>56.0</td>\n", |
||
| 775 | " <td>0.43</td>\n", |
||
| 776 | " <td>1.0</td>\n", |
||
| 777 | " <td>...</td>\n", |
||
| 778 | " <td>-0.0781</td>\n", |
||
| 779 | " <td>-0.2368</td>\n", |
||
| 780 | " <td>0.5056</td>\n", |
||
| 781 | " <td>-0.2771</td>\n", |
||
| 782 | " <td>0.1067</td>\n", |
||
| 783 | " <td>-0.0153</td>\n", |
||
| 784 | " <td>-0.2546</td>\n", |
||
| 785 | " <td>-0.4205</td>\n", |
||
| 786 | " <td>-0.3773</td>\n", |
||
| 787 | " <td>0.0551</td>\n", |
||
| 788 | " </tr>\n", |
||
| 789 | " <tr>\n", |
||
| 790 | " <th>512</th>\n", |
||
| 791 | " <td>512</td>\n", |
||
| 792 | " <td>512</td>\n", |
||
| 793 | " <td>TCGA-NJ-A55R</td>\n", |
||
| 794 | " <td>TCGA-NJ-A55R-01Z-00-DX1.2E2B3642-4E1C-47DB-AF7...</td>\n", |
||
| 795 | " <td>NJ</td>\n", |
||
| 796 | " <td>0.0</td>\n", |
||
| 797 | " <td>LUAD</td>\n", |
||
| 798 | " <td>67.0</td>\n", |
||
| 799 | " <td>19.81</td>\n", |
||
| 800 | " <td>1.0</td>\n", |
||
| 801 | " <td>...</td>\n", |
||
| 802 | " <td>6.1880</td>\n", |
||
| 803 | " <td>0.2405</td>\n", |
||
| 804 | " <td>0.0751</td>\n", |
||
| 805 | " <td>1.9723</td>\n", |
||
| 806 | " <td>0.6093</td>\n", |
||
| 807 | " <td>0.6135</td>\n", |
||
| 808 | " <td>1.7846</td>\n", |
||
| 809 | " <td>0.0588</td>\n", |
||
| 810 | " <td>-0.1157</td>\n", |
||
| 811 | " <td>1.2831</td>\n", |
||
| 812 | " </tr>\n", |
||
| 813 | " <tr>\n", |
||
| 814 | " <th>513</th>\n", |
||
| 815 | " <td>513</td>\n", |
||
| 816 | " <td>513</td>\n", |
||
| 817 | " <td>TCGA-NJ-A7XG</td>\n", |
||
| 818 | " <td>TCGA-NJ-A7XG-01Z-00-DX1.4A876254-653C-410B-A36...</td>\n", |
||
| 819 | " <td>NJ</td>\n", |
||
| 820 | " <td>0.0</td>\n", |
||
| 821 | " <td>LUAD</td>\n", |
||
| 822 | " <td>49.0</td>\n", |
||
| 823 | " <td>20.27</td>\n", |
||
| 824 | " <td>1.0</td>\n", |
||
| 825 | " <td>...</td>\n", |
||
| 826 | " <td>-0.1238</td>\n", |
||
| 827 | " <td>-0.0041</td>\n", |
||
| 828 | " <td>-0.8129</td>\n", |
||
| 829 | " <td>-0.4409</td>\n", |
||
| 830 | " <td>0.6778</td>\n", |
||
| 831 | " <td>-0.5506</td>\n", |
||
| 832 | " <td>1.4350</td>\n", |
||
| 833 | " <td>-1.5823</td>\n", |
||
| 834 | " <td>-1.3015</td>\n", |
||
| 835 | " <td>0.4371</td>\n", |
||
| 836 | " </tr>\n", |
||
| 837 | " <tr>\n", |
||
| 838 | " <th>514</th>\n", |
||
| 839 | " <td>514</td>\n", |
||
| 840 | " <td>514</td>\n", |
||
| 841 | " <td>TCGA-O1-A52J</td>\n", |
||
| 842 | " <td>TCGA-O1-A52J-01Z-00-DX1.26F6ECCA-D614-4950-98E...</td>\n", |
||
| 843 | " <td>O1</td>\n", |
||
| 844 | " <td>1.0</td>\n", |
||
| 845 | " <td>LUAD</td>\n", |
||
| 846 | " <td>74.0</td>\n", |
||
| 847 | " <td>59.07</td>\n", |
||
| 848 | " <td>0.0</td>\n", |
||
| 849 | " <td>...</td>\n", |
||
| 850 | " <td>-0.1238</td>\n", |
||
| 851 | " <td>-0.1263</td>\n", |
||
| 852 | " <td>0.8472</td>\n", |
||
| 853 | " <td>-0.3943</td>\n", |
||
| 854 | " <td>-0.7671</td>\n", |
||
| 855 | " <td>-1.1313</td>\n", |
||
| 856 | " <td>-0.9671</td>\n", |
||
| 857 | " <td>4.2234</td>\n", |
||
| 858 | " <td>0.9716</td>\n", |
||
| 859 | " <td>0.6699</td>\n", |
||
| 860 | " </tr>\n", |
||
| 861 | " <tr>\n", |
||
| 862 | " <th>515</th>\n", |
||
| 863 | " <td>515</td>\n", |
||
| 864 | " <td>515</td>\n", |
||
| 865 | " <td>TCGA-S2-AA1A</td>\n", |
||
| 866 | " <td>TCGA-S2-AA1A-01Z-00-DX1.4B5D5FAE-8305-4D2D-B24...</td>\n", |
||
| 867 | " <td>S2</td>\n", |
||
| 868 | " <td>1.0</td>\n", |
||
| 869 | " <td>LUAD</td>\n", |
||
| 870 | " <td>68.0</td>\n", |
||
| 871 | " <td>16.85</td>\n", |
||
| 872 | " <td>1.0</td>\n", |
||
| 873 | " <td>...</td>\n", |
||
| 874 | " <td>-0.1238</td>\n", |
||
| 875 | " <td>0.5292</td>\n", |
||
| 876 | " <td>-0.8343</td>\n", |
||
| 877 | " <td>0.7741</td>\n", |
||
| 878 | " <td>-0.6405</td>\n", |
||
| 879 | " <td>-0.3901</td>\n", |
||
| 880 | " <td>0.0245</td>\n", |
||
| 881 | " <td>0.5245</td>\n", |
||
| 882 | " <td>-0.1738</td>\n", |
||
| 883 | " <td>2.4043</td>\n", |
||
| 884 | " </tr>\n", |
||
| 885 | " </tbody>\n", |
||
| 886 | "</table>\n", |
||
| 887 | "<p>516 rows × 3106 columns</p>\n", |
||
| 888 | "</div>" |
||
| 889 | ], |
||
| 890 | "text/plain": [ |
||
| 891 | " Unnamed: 0 Unnamed: 0.1 case_id \\\n", |
||
| 892 | "0 0 0 TCGA-05-4249 \n", |
||
| 893 | "1 1 1 TCGA-05-4250 \n", |
||
| 894 | "2 2 2 TCGA-05-4382 \n", |
||
| 895 | "3 3 3 TCGA-05-4384 \n", |
||
| 896 | "4 4 4 TCGA-05-4390 \n", |
||
| 897 | ".. ... ... ... \n", |
||
| 898 | "511 511 511 TCGA-NJ-A55O \n", |
||
| 899 | "512 512 512 TCGA-NJ-A55R \n", |
||
| 900 | "513 513 513 TCGA-NJ-A7XG \n", |
||
| 901 | "514 514 514 TCGA-O1-A52J \n", |
||
| 902 | "515 515 515 TCGA-S2-AA1A \n", |
||
| 903 | "\n", |
||
| 904 | " slide_id site is_female \\\n", |
||
| 905 | "0 TCGA-05-4249-01Z-00-DX1.9fce0297-cc19-4c04-872... 5 0.0 \n", |
||
| 906 | "1 TCGA-05-4250-01Z-00-DX1.90f67fdf-dff9-46ca-af7... 5 1.0 \n", |
||
| 907 | "2 TCGA-05-4382-01Z-00-DX1.76b49a4c-dbbb-48b0-b67... 5 0.0 \n", |
||
| 908 | "3 TCGA-05-4384-01Z-00-DX1.CA68BF29-BBE3-4C8E-B48... 5 0.0 \n", |
||
| 909 | "4 TCGA-05-4390-01Z-00-DX1.858E64DF-DD3E-4F43-B7C... 5 1.0 \n", |
||
| 910 | ".. ... ... ... \n", |
||
| 911 | "511 TCGA-NJ-A55O-01Z-00-DX1.8E23C821-B8BB-4D89-9E3... NJ 1.0 \n", |
||
| 912 | "512 TCGA-NJ-A55R-01Z-00-DX1.2E2B3642-4E1C-47DB-AF7... NJ 0.0 \n", |
||
| 913 | "513 TCGA-NJ-A7XG-01Z-00-DX1.4A876254-653C-410B-A36... NJ 0.0 \n", |
||
| 914 | "514 TCGA-O1-A52J-01Z-00-DX1.26F6ECCA-D614-4950-98E... O1 1.0 \n", |
||
| 915 | "515 TCGA-S2-AA1A-01Z-00-DX1.4B5D5FAE-8305-4D2D-B24... S2 1.0 \n", |
||
| 916 | "\n", |
||
| 917 | " oncotree_code age survival_months censorship ... ZSCAN10_rnaseq \\\n", |
||
| 918 | "0 LUAD 67.0 50.03 1.0 ... -0.1238 \n", |
||
| 919 | "1 LUAD 79.0 3.98 0.0 ... -0.1238 \n", |
||
| 920 | "2 LUAD 68.0 19.94 1.0 ... 0.3265 \n", |
||
| 921 | "3 LUAD 66.0 13.99 1.0 ... -0.1238 \n", |
||
| 922 | "4 LUAD 58.0 36.99 1.0 ... -0.1238 \n", |
||
| 923 | ".. ... ... ... ... ... ... \n", |
||
| 924 | "511 LUAD 56.0 0.43 1.0 ... -0.0781 \n", |
||
| 925 | "512 LUAD 67.0 19.81 1.0 ... 6.1880 \n", |
||
| 926 | "513 LUAD 49.0 20.27 1.0 ... -0.1238 \n", |
||
| 927 | "514 LUAD 74.0 59.07 0.0 ... -0.1238 \n", |
||
| 928 | "515 LUAD 68.0 16.85 1.0 ... -0.1238 \n", |
||
| 929 | "\n", |
||
| 930 | " ZSCAN12_rnaseq ZSCAN20_rnaseq ZSCAN21_rnaseq ZSCAN22_rnaseq \\\n", |
||
| 931 | "0 0.7530 0.6552 -1.0013 0.2353 \n", |
||
| 932 | "1 0.4810 -0.8255 0.2825 -1.2502 \n", |
||
| 933 | "2 0.4462 1.1847 0.8765 -0.7999 \n", |
||
| 934 | "3 -0.0369 0.5766 0.0083 0.1344 \n", |
||
| 935 | "4 0.4751 1.2404 0.6932 -0.2792 \n", |
||
| 936 | ".. ... ... ... ... \n", |
||
| 937 | "511 -0.2368 0.5056 -0.2771 0.1067 \n", |
||
| 938 | "512 0.2405 0.0751 1.9723 0.6093 \n", |
||
| 939 | "513 -0.0041 -0.8129 -0.4409 0.6778 \n", |
||
| 940 | "514 -0.1263 0.8472 -0.3943 -0.7671 \n", |
||
| 941 | "515 0.5292 -0.8343 0.7741 -0.6405 \n", |
||
| 942 | "\n", |
||
| 943 | " ZSCAN2_rnaseq ZSCAN9_rnaseq ZXDA_rnaseq ZXDB_rnaseq ZXDC_rnaseq \n", |
||
| 944 | "0 2.6532 1.1103 0.6149 0.5725 0.2889 \n", |
||
| 945 | "1 -0.9024 -0.1472 0.5118 -0.1673 -0.8006 \n", |
||
| 946 | "2 1.7566 1.1757 -0.4399 -0.2751 -0.4668 \n", |
||
| 947 | "3 0.8299 0.6599 1.4844 0.9748 0.7481 \n", |
||
| 948 | "4 2.1326 0.1621 -0.0462 1.8418 -0.9922 \n", |
||
| 949 | ".. ... ... ... ... ... \n", |
||
| 950 | "511 -0.0153 -0.2546 -0.4205 -0.3773 0.0551 \n", |
||
| 951 | "512 0.6135 1.7846 0.0588 -0.1157 1.2831 \n", |
||
| 952 | "513 -0.5506 1.4350 -1.5823 -1.3015 0.4371 \n", |
||
| 953 | "514 -1.1313 -0.9671 4.2234 0.9716 0.6699 \n", |
||
| 954 | "515 -0.3901 0.0245 0.5245 -0.1738 2.4043 \n", |
||
| 955 | "\n", |
||
| 956 | "[516 rows x 3106 columns]" |
||
| 957 | ] |
||
| 958 | }, |
||
| 959 | "execution_count": 81, |
||
| 960 | "metadata": {}, |
||
| 961 | "output_type": "execute_result" |
||
| 962 | } |
||
| 963 | ], |
||
| 964 | "source": [ |
||
| 965 | "fname = '../dataset_csv/tcga_luad_all_clean.csv.zip'\n", |
||
| 966 | "slide_df = pd.read_csv(fname)\n", |
||
| 967 | "slide_df" |
||
| 968 | ] |
||
| 969 | }, |
||
| 970 | { |
||
| 971 | "cell_type": "code", |
||
| 972 | "execution_count": 79, |
||
| 973 | "metadata": {}, |
||
| 974 | "outputs": [], |
||
| 975 | "source": [ |
||
| 976 | "fname = '../dataset_csv/tcga_luad_all_clean.csv.zip'\n", |
||
| 977 | "slide_df = pd.read_csv(fname)\n", |
||
| 978 | "omic_overlap = np.concatenate([omic_from_signatures+mode for mode in ['_mut', '_cnv', '_rnaseq']])\n", |
||
| 979 | "omic_overlap = sorted(series_intersection(omic_overlap, slide_df.columns))\n", |
||
| 980 | "slide_df[list(slide_df.columns[:9]) + omic_overlap].to_csv('../dataset_csv_sig/%s' % fname)" |
||
| 981 | ] |
||
| 982 | }, |
||
| 983 | { |
||
| 984 | "cell_type": "code", |
||
| 985 | "execution_count": 54, |
||
| 986 | "metadata": {}, |
||
| 987 | "outputs": [ |
||
| 988 | { |
||
| 989 | "name": "stderr", |
||
| 990 | "output_type": "stream", |
||
| 991 | "text": [ |
||
| 992 | "/home/mahmoodlab/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3071: DtypeWarning: Columns (3) have mixed types.Specify dtype option on import or set low_memory=False.\n", |
||
| 993 | " has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n" |
||
| 994 | ] |
||
| 995 | } |
||
| 996 | ], |
||
| 997 | "source": [ |
||
| 998 | "for fname in os.listdir('./'):\n", |
||
| 999 | " if fname.endswith('.csv.zip'):\n", |
||
| 1000 | " slide_df = pd.read_csv(fname)" |
||
| 1001 | ] |
||
| 1002 | }, |
||
| 1003 | { |
||
| 1004 | "cell_type": "code", |
||
| 1005 | "execution_count": 55, |
||
| 1006 | "metadata": {}, |
||
| 1007 | "outputs": [ |
||
| 1008 | { |
||
| 1009 | "data": { |
||
| 1010 | "text/plain": [ |
||
| 1011 | "'tcga_luad_all_clean.csv.zip'" |
||
| 1012 | ] |
||
| 1013 | }, |
||
| 1014 | "execution_count": 55, |
||
| 1015 | "metadata": {}, |
||
| 1016 | "output_type": "execute_result" |
||
| 1017 | } |
||
| 1018 | ], |
||
| 1019 | "source": [ |
||
| 1020 | "fname" |
||
| 1021 | ] |
||
| 1022 | }, |
||
| 1023 | { |
||
| 1024 | "cell_type": "code", |
||
| 1025 | "execution_count": null, |
||
| 1026 | "metadata": {}, |
||
| 1027 | "outputs": [], |
||
| 1028 | "source": [] |
||
| 1029 | }, |
||
| 1030 | { |
||
| 1031 | "cell_type": "code", |
||
| 1032 | "execution_count": null, |
||
| 1033 | "metadata": {}, |
||
| 1034 | "outputs": [], |
||
| 1035 | "source": [] |
||
| 1036 | }, |
||
| 1037 | { |
||
| 1038 | "cell_type": "code", |
||
| 1039 | "execution_count": null, |
||
| 1040 | "metadata": {}, |
||
| 1041 | "outputs": [], |
||
| 1042 | "source": [] |
||
| 1043 | }, |
||
| 1044 | { |
||
| 1045 | "cell_type": "code", |
||
| 1046 | "execution_count": 15, |
||
| 1047 | "metadata": {}, |
||
| 1048 | "outputs": [ |
||
| 1049 | { |
||
| 1050 | "name": "stderr", |
||
| 1051 | "output_type": "stream", |
||
| 1052 | "text": [ |
||
| 1053 | "/home/mahmoodlab/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3071: DtypeWarning: Columns (2) have mixed types.Specify dtype option on import or set low_memory=False.\n", |
||
| 1054 | " has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n" |
||
| 1055 | ] |
||
| 1056 | } |
||
| 1057 | ], |
||
| 1058 | "source": [ |
||
| 1059 | "\n", |
||
| 1060 | "\n", |
||
| 1061 | "omic_from_signatures = []\n", |
||
| 1062 | "for col in signatures.columns:\n", |
||
| 1063 | " omic = signatures[col].dropna().unique()\n", |
||
| 1064 | " omic_from_signatures.append(omic)\n", |
||
| 1065 | "\n", |
||
| 1066 | "omic_from_signatures = np.concatenate(omic_from_signatures)\n", |
||
| 1067 | "\n", |
||
| 1068 | "def series_intersection(s1, s2):\n", |
||
| 1069 | " return pd.Series(list(set(s1) & set(s2)))\n", |
||
| 1070 | "\n", |
||
| 1071 | "signatures = pd.read_csv('./signatures.csv')\n", |
||
| 1072 | "slide_df = pd.read_csv('./tcga_gbmlgg_all_clean.csv.zip')\n", |
||
| 1073 | "rnaseq_overlap = np.concatenate([omic_from_signatures+mode for mode in ['_rnaseq']])\n", |
||
| 1074 | "rnaseq_overlap = sorted(series_intersection(rnaseq_overlap, slide_df.columns))\n", |
||
| 1075 | "genomics_mut_cnv = list(slide_df.columns[slide_df.columns.str.contains('_mut|_cnv')])" |
||
| 1076 | ] |
||
| 1077 | }, |
||
| 1078 | { |
||
| 1079 | "cell_type": "code", |
||
| 1080 | "execution_count": 16, |
||
| 1081 | "metadata": {}, |
||
| 1082 | "outputs": [], |
||
| 1083 | "source": [ |
||
| 1084 | "_ = slide_df[list(slide_df.columns[:9]) + rnaseq_overlap + genomics_mut_cnv]" |
||
| 1085 | ] |
||
| 1086 | }, |
||
| 1087 | { |
||
| 1088 | "cell_type": "code", |
||
| 1089 | "execution_count": 17, |
||
| 1090 | "metadata": {}, |
||
| 1091 | "outputs": [ |
||
| 1092 | { |
||
| 1093 | "data": { |
||
| 1094 | "text/html": [ |
||
| 1095 | "<div>\n", |
||
| 1096 | "<style scoped>\n", |
||
| 1097 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 1098 | " vertical-align: middle;\n", |
||
| 1099 | " }\n", |
||
| 1100 | "\n", |
||
| 1101 | " .dataframe tbody tr th {\n", |
||
| 1102 | " vertical-align: top;\n", |
||
| 1103 | " }\n", |
||
| 1104 | "\n", |
||
| 1105 | " .dataframe thead th {\n", |
||
| 1106 | " text-align: right;\n", |
||
| 1107 | " }\n", |
||
| 1108 | "</style>\n", |
||
| 1109 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 1110 | " <thead>\n", |
||
| 1111 | " <tr style=\"text-align: right;\">\n", |
||
| 1112 | " <th></th>\n", |
||
| 1113 | " <th>case_id</th>\n", |
||
| 1114 | " <th>slide_id</th>\n", |
||
| 1115 | " <th>site</th>\n", |
||
| 1116 | " <th>is_female</th>\n", |
||
| 1117 | " <th>oncotree_code</th>\n", |
||
| 1118 | " <th>age</th>\n", |
||
| 1119 | " <th>survival_months</th>\n", |
||
| 1120 | " <th>censorship</th>\n", |
||
| 1121 | " <th>train</th>\n", |
||
| 1122 | " <th>AAK1_rnaseq</th>\n", |
||
| 1123 | " <th>...</th>\n", |
||
| 1124 | " <th>AGAP2_cnv</th>\n", |
||
| 1125 | " <th>TSPAN31_cnv</th>\n", |
||
| 1126 | " <th>CDK4_cnv</th>\n", |
||
| 1127 | " <th>MARCH9_cnv</th>\n", |
||
| 1128 | " <th>CYP27B1_cnv</th>\n", |
||
| 1129 | " <th>METTL1_cnv</th>\n", |
||
| 1130 | " <th>TSFM_cnv</th>\n", |
||
| 1131 | " <th>AVIL_cnv</th>\n", |
||
| 1132 | " <th>CTDSP2_cnv</th>\n", |
||
| 1133 | " <th>RN7SKP65_cnv</th>\n", |
||
| 1134 | " </tr>\n", |
||
| 1135 | " </thead>\n", |
||
| 1136 | " <tbody>\n", |
||
| 1137 | " <tr>\n", |
||
| 1138 | " <th>0</th>\n", |
||
| 1139 | " <td>TCGA-02-0047</td>\n", |
||
| 1140 | " <td>TCGA-02-0047-01Z-00-DX1.4755D138-5842-4159-848...</td>\n", |
||
| 1141 | " <td>2</td>\n", |
||
| 1142 | " <td>0.0</td>\n", |
||
| 1143 | " <td>GBM</td>\n", |
||
| 1144 | " <td>78.0</td>\n", |
||
| 1145 | " <td>14.72</td>\n", |
||
| 1146 | " <td>0.0</td>\n", |
||
| 1147 | " <td>1.0</td>\n", |
||
| 1148 | " <td>1.5517</td>\n", |
||
| 1149 | " <td>...</td>\n", |
||
| 1150 | " <td>0</td>\n", |
||
| 1151 | " <td>0</td>\n", |
||
| 1152 | " <td>0</td>\n", |
||
| 1153 | " <td>0</td>\n", |
||
| 1154 | " <td>0</td>\n", |
||
| 1155 | " <td>0</td>\n", |
||
| 1156 | " <td>0</td>\n", |
||
| 1157 | " <td>0</td>\n", |
||
| 1158 | " <td>0</td>\n", |
||
| 1159 | " <td>0</td>\n", |
||
| 1160 | " </tr>\n", |
||
| 1161 | " <tr>\n", |
||
| 1162 | " <th>1</th>\n", |
||
| 1163 | " <td>TCGA-06-0125</td>\n", |
||
| 1164 | " <td>TCGA-06-0125-01Z-00-DX1.8e0915b2-8dc3-4753-806...</td>\n", |
||
| 1165 | " <td>6</td>\n", |
||
| 1166 | " <td>1.0</td>\n", |
||
| 1167 | " <td>GBM</td>\n", |
||
| 1168 | " <td>63.0</td>\n", |
||
| 1169 | " <td>47.57</td>\n", |
||
| 1170 | " <td>0.0</td>\n", |
||
| 1171 | " <td>1.0</td>\n", |
||
| 1172 | " <td>0.5557</td>\n", |
||
| 1173 | " <td>...</td>\n", |
||
| 1174 | " <td>0</td>\n", |
||
| 1175 | " <td>0</td>\n", |
||
| 1176 | " <td>0</td>\n", |
||
| 1177 | " <td>0</td>\n", |
||
| 1178 | " <td>0</td>\n", |
||
| 1179 | " <td>0</td>\n", |
||
| 1180 | " <td>0</td>\n", |
||
| 1181 | " <td>0</td>\n", |
||
| 1182 | " <td>0</td>\n", |
||
| 1183 | " <td>0</td>\n", |
||
| 1184 | " </tr>\n", |
||
| 1185 | " <tr>\n", |
||
| 1186 | " <th>2</th>\n", |
||
| 1187 | " <td>TCGA-06-0125</td>\n", |
||
| 1188 | " <td>TCGA-06-0125-01Z-00-DX2.4f9cef92-2bdb-480d-870...</td>\n", |
||
| 1189 | " <td>6</td>\n", |
||
| 1190 | " <td>1.0</td>\n", |
||
| 1191 | " <td>GBM</td>\n", |
||
| 1192 | " <td>63.0</td>\n", |
||
| 1193 | " <td>47.57</td>\n", |
||
| 1194 | " <td>0.0</td>\n", |
||
| 1195 | " <td>1.0</td>\n", |
||
| 1196 | " <td>0.5557</td>\n", |
||
| 1197 | " <td>...</td>\n", |
||
| 1198 | " <td>0</td>\n", |
||
| 1199 | " <td>0</td>\n", |
||
| 1200 | " <td>0</td>\n", |
||
| 1201 | " <td>0</td>\n", |
||
| 1202 | " <td>0</td>\n", |
||
| 1203 | " <td>0</td>\n", |
||
| 1204 | " <td>0</td>\n", |
||
| 1205 | " <td>0</td>\n", |
||
| 1206 | " <td>0</td>\n", |
||
| 1207 | " <td>0</td>\n", |
||
| 1208 | " </tr>\n", |
||
| 1209 | " <tr>\n", |
||
| 1210 | " <th>3</th>\n", |
||
| 1211 | " <td>TCGA-06-0129</td>\n", |
||
| 1212 | " <td>TCGA-06-0129-01Z-00-DX1.b7bddf7d-f39e-45e7-a78...</td>\n", |
||
| 1213 | " <td>6</td>\n", |
||
| 1214 | " <td>0.0</td>\n", |
||
| 1215 | " <td>GBM</td>\n", |
||
| 1216 | " <td>30.0</td>\n", |
||
| 1217 | " <td>33.64</td>\n", |
||
| 1218 | " <td>0.0</td>\n", |
||
| 1219 | " <td>1.0</td>\n", |
||
| 1220 | " <td>0.6442</td>\n", |
||
| 1221 | " <td>...</td>\n", |
||
| 1222 | " <td>2</td>\n", |
||
| 1223 | " <td>2</td>\n", |
||
| 1224 | " <td>2</td>\n", |
||
| 1225 | " <td>2</td>\n", |
||
| 1226 | " <td>2</td>\n", |
||
| 1227 | " <td>2</td>\n", |
||
| 1228 | " <td>2</td>\n", |
||
| 1229 | " <td>2</td>\n", |
||
| 1230 | " <td>2</td>\n", |
||
| 1231 | " <td>2</td>\n", |
||
| 1232 | " </tr>\n", |
||
| 1233 | " <tr>\n", |
||
| 1234 | " <th>4</th>\n", |
||
| 1235 | " <td>TCGA-06-0129</td>\n", |
||
| 1236 | " <td>TCGA-06-0129-01Z-00-DX2.1ea78b46-1dc7-44d8-81b...</td>\n", |
||
| 1237 | " <td>6</td>\n", |
||
| 1238 | " <td>0.0</td>\n", |
||
| 1239 | " <td>GBM</td>\n", |
||
| 1240 | " <td>30.0</td>\n", |
||
| 1241 | " <td>33.64</td>\n", |
||
| 1242 | " <td>0.0</td>\n", |
||
| 1243 | " <td>1.0</td>\n", |
||
| 1244 | " <td>0.6442</td>\n", |
||
| 1245 | " <td>...</td>\n", |
||
| 1246 | " <td>2</td>\n", |
||
| 1247 | " <td>2</td>\n", |
||
| 1248 | " <td>2</td>\n", |
||
| 1249 | " <td>2</td>\n", |
||
| 1250 | " <td>2</td>\n", |
||
| 1251 | " <td>2</td>\n", |
||
| 1252 | " <td>2</td>\n", |
||
| 1253 | " <td>2</td>\n", |
||
| 1254 | " <td>2</td>\n", |
||
| 1255 | " <td>2</td>\n", |
||
| 1256 | " </tr>\n", |
||
| 1257 | " <tr>\n", |
||
| 1258 | " <th>...</th>\n", |
||
| 1259 | " <td>...</td>\n", |
||
| 1260 | " <td>...</td>\n", |
||
| 1261 | " <td>...</td>\n", |
||
| 1262 | " <td>...</td>\n", |
||
| 1263 | " <td>...</td>\n", |
||
| 1264 | " <td>...</td>\n", |
||
| 1265 | " <td>...</td>\n", |
||
| 1266 | " <td>...</td>\n", |
||
| 1267 | " <td>...</td>\n", |
||
| 1268 | " <td>...</td>\n", |
||
| 1269 | " <td>...</td>\n", |
||
| 1270 | " <td>...</td>\n", |
||
| 1271 | " <td>...</td>\n", |
||
| 1272 | " <td>...</td>\n", |
||
| 1273 | " <td>...</td>\n", |
||
| 1274 | " <td>...</td>\n", |
||
| 1275 | " <td>...</td>\n", |
||
| 1276 | " <td>...</td>\n", |
||
| 1277 | " <td>...</td>\n", |
||
| 1278 | " <td>...</td>\n", |
||
| 1279 | " <td>...</td>\n", |
||
| 1280 | " </tr>\n", |
||
| 1281 | " <tr>\n", |
||
| 1282 | " <th>1037</th>\n", |
||
| 1283 | " <td>TCGA-WY-A85A</td>\n", |
||
| 1284 | " <td>TCGA-WY-A85A-01Z-00-DX1.CB302B89-F89A-40FD-A7D...</td>\n", |
||
| 1285 | " <td>WY</td>\n", |
||
| 1286 | " <td>0.0</td>\n", |
||
| 1287 | " <td>ASTR</td>\n", |
||
| 1288 | " <td>20.0</td>\n", |
||
| 1289 | " <td>43.36</td>\n", |
||
| 1290 | " <td>1.0</td>\n", |
||
| 1291 | " <td>1.0</td>\n", |
||
| 1292 | " <td>-0.3841</td>\n", |
||
| 1293 | " <td>...</td>\n", |
||
| 1294 | " <td>0</td>\n", |
||
| 1295 | " <td>0</td>\n", |
||
| 1296 | " <td>0</td>\n", |
||
| 1297 | " <td>0</td>\n", |
||
| 1298 | " <td>0</td>\n", |
||
| 1299 | " <td>0</td>\n", |
||
| 1300 | " <td>0</td>\n", |
||
| 1301 | " <td>0</td>\n", |
||
| 1302 | " <td>0</td>\n", |
||
| 1303 | " <td>0</td>\n", |
||
| 1304 | " </tr>\n", |
||
| 1305 | " <tr>\n", |
||
| 1306 | " <th>1038</th>\n", |
||
| 1307 | " <td>TCGA-WY-A85B</td>\n", |
||
| 1308 | " <td>TCGA-WY-A85B-01Z-00-DX1.1E4B796A-A1E3-45F9-807...</td>\n", |
||
| 1309 | " <td>WY</td>\n", |
||
| 1310 | " <td>0.0</td>\n", |
||
| 1311 | " <td>ASTR</td>\n", |
||
| 1312 | " <td>24.0</td>\n", |
||
| 1313 | " <td>45.76</td>\n", |
||
| 1314 | " <td>1.0</td>\n", |
||
| 1315 | " <td>1.0</td>\n", |
||
| 1316 | " <td>-0.4479</td>\n", |
||
| 1317 | " <td>...</td>\n", |
||
| 1318 | " <td>-1</td>\n", |
||
| 1319 | " <td>-1</td>\n", |
||
| 1320 | " <td>-1</td>\n", |
||
| 1321 | " <td>-1</td>\n", |
||
| 1322 | " <td>-1</td>\n", |
||
| 1323 | " <td>-1</td>\n", |
||
| 1324 | " <td>-1</td>\n", |
||
| 1325 | " <td>-1</td>\n", |
||
| 1326 | " <td>-1</td>\n", |
||
| 1327 | " <td>-1</td>\n", |
||
| 1328 | " </tr>\n", |
||
| 1329 | " <tr>\n", |
||
| 1330 | " <th>1039</th>\n", |
||
| 1331 | " <td>TCGA-WY-A85C</td>\n", |
||
| 1332 | " <td>TCGA-WY-A85C-01Z-00-DX1.E0A6429A-91B3-4FFE-9FF...</td>\n", |
||
| 1333 | " <td>WY</td>\n", |
||
| 1334 | " <td>0.0</td>\n", |
||
| 1335 | " <td>ASTR</td>\n", |
||
| 1336 | " <td>36.0</td>\n", |
||
| 1337 | " <td>46.85</td>\n", |
||
| 1338 | " <td>1.0</td>\n", |
||
| 1339 | " <td>1.0</td>\n", |
||
| 1340 | " <td>-0.2472</td>\n", |
||
| 1341 | " <td>...</td>\n", |
||
| 1342 | " <td>0</td>\n", |
||
| 1343 | " <td>0</td>\n", |
||
| 1344 | " <td>0</td>\n", |
||
| 1345 | " <td>0</td>\n", |
||
| 1346 | " <td>0</td>\n", |
||
| 1347 | " <td>0</td>\n", |
||
| 1348 | " <td>0</td>\n", |
||
| 1349 | " <td>0</td>\n", |
||
| 1350 | " <td>0</td>\n", |
||
| 1351 | " <td>0</td>\n", |
||
| 1352 | " </tr>\n", |
||
| 1353 | " <tr>\n", |
||
| 1354 | " <th>1040</th>\n", |
||
| 1355 | " <td>TCGA-WY-A85D</td>\n", |
||
| 1356 | " <td>TCGA-WY-A85D-01Z-00-DX1.FB8C252B-7A88-4B14-B3C...</td>\n", |
||
| 1357 | " <td>WY</td>\n", |
||
| 1358 | " <td>0.0</td>\n", |
||
| 1359 | " <td>OAST</td>\n", |
||
| 1360 | " <td>60.0</td>\n", |
||
| 1361 | " <td>37.68</td>\n", |
||
| 1362 | " <td>1.0</td>\n", |
||
| 1363 | " <td>1.0</td>\n", |
||
| 1364 | " <td>-0.5892</td>\n", |
||
| 1365 | " <td>...</td>\n", |
||
| 1366 | " <td>0</td>\n", |
||
| 1367 | " <td>0</td>\n", |
||
| 1368 | " <td>0</td>\n", |
||
| 1369 | " <td>0</td>\n", |
||
| 1370 | " <td>0</td>\n", |
||
| 1371 | " <td>0</td>\n", |
||
| 1372 | " <td>0</td>\n", |
||
| 1373 | " <td>0</td>\n", |
||
| 1374 | " <td>0</td>\n", |
||
| 1375 | " <td>0</td>\n", |
||
| 1376 | " </tr>\n", |
||
| 1377 | " <tr>\n", |
||
| 1378 | " <th>1041</th>\n", |
||
| 1379 | " <td>TCGA-WY-A85E</td>\n", |
||
| 1380 | " <td>TCGA-WY-A85E-01Z-00-DX1.AA7A4C1F-99AA-490D-B6D...</td>\n", |
||
| 1381 | " <td>WY</td>\n", |
||
| 1382 | " <td>1.0</td>\n", |
||
| 1383 | " <td>OAST</td>\n", |
||
| 1384 | " <td>48.0</td>\n", |
||
| 1385 | " <td>20.80</td>\n", |
||
| 1386 | " <td>1.0</td>\n", |
||
| 1387 | " <td>1.0</td>\n", |
||
| 1388 | " <td>-0.1087</td>\n", |
||
| 1389 | " <td>...</td>\n", |
||
| 1390 | " <td>0</td>\n", |
||
| 1391 | " <td>0</td>\n", |
||
| 1392 | " <td>0</td>\n", |
||
| 1393 | " <td>0</td>\n", |
||
| 1394 | " <td>0</td>\n", |
||
| 1395 | " <td>0</td>\n", |
||
| 1396 | " <td>0</td>\n", |
||
| 1397 | " <td>0</td>\n", |
||
| 1398 | " <td>0</td>\n", |
||
| 1399 | " <td>0</td>\n", |
||
| 1400 | " </tr>\n", |
||
| 1401 | " </tbody>\n", |
||
| 1402 | "</table>\n", |
||
| 1403 | "<p>1042 rows × 2891 columns</p>\n", |
||
| 1404 | "</div>" |
||
| 1405 | ], |
||
| 1406 | "text/plain": [ |
||
| 1407 | " case_id slide_id site \\\n", |
||
| 1408 | "0 TCGA-02-0047 TCGA-02-0047-01Z-00-DX1.4755D138-5842-4159-848... 2 \n", |
||
| 1409 | "1 TCGA-06-0125 TCGA-06-0125-01Z-00-DX1.8e0915b2-8dc3-4753-806... 6 \n", |
||
| 1410 | "2 TCGA-06-0125 TCGA-06-0125-01Z-00-DX2.4f9cef92-2bdb-480d-870... 6 \n", |
||
| 1411 | "3 TCGA-06-0129 TCGA-06-0129-01Z-00-DX1.b7bddf7d-f39e-45e7-a78... 6 \n", |
||
| 1412 | "4 TCGA-06-0129 TCGA-06-0129-01Z-00-DX2.1ea78b46-1dc7-44d8-81b... 6 \n", |
||
| 1413 | "... ... ... ... \n", |
||
| 1414 | "1037 TCGA-WY-A85A TCGA-WY-A85A-01Z-00-DX1.CB302B89-F89A-40FD-A7D... WY \n", |
||
| 1415 | "1038 TCGA-WY-A85B TCGA-WY-A85B-01Z-00-DX1.1E4B796A-A1E3-45F9-807... WY \n", |
||
| 1416 | "1039 TCGA-WY-A85C TCGA-WY-A85C-01Z-00-DX1.E0A6429A-91B3-4FFE-9FF... WY \n", |
||
| 1417 | "1040 TCGA-WY-A85D TCGA-WY-A85D-01Z-00-DX1.FB8C252B-7A88-4B14-B3C... WY \n", |
||
| 1418 | "1041 TCGA-WY-A85E TCGA-WY-A85E-01Z-00-DX1.AA7A4C1F-99AA-490D-B6D... WY \n", |
||
| 1419 | "\n", |
||
| 1420 | " is_female oncotree_code age survival_months censorship train \\\n", |
||
| 1421 | "0 0.0 GBM 78.0 14.72 0.0 1.0 \n", |
||
| 1422 | "1 1.0 GBM 63.0 47.57 0.0 1.0 \n", |
||
| 1423 | "2 1.0 GBM 63.0 47.57 0.0 1.0 \n", |
||
| 1424 | "3 0.0 GBM 30.0 33.64 0.0 1.0 \n", |
||
| 1425 | "4 0.0 GBM 30.0 33.64 0.0 1.0 \n", |
||
| 1426 | "... ... ... ... ... ... ... \n", |
||
| 1427 | "1037 0.0 ASTR 20.0 43.36 1.0 1.0 \n", |
||
| 1428 | "1038 0.0 ASTR 24.0 45.76 1.0 1.0 \n", |
||
| 1429 | "1039 0.0 ASTR 36.0 46.85 1.0 1.0 \n", |
||
| 1430 | "1040 0.0 OAST 60.0 37.68 1.0 1.0 \n", |
||
| 1431 | "1041 1.0 OAST 48.0 20.80 1.0 1.0 \n", |
||
| 1432 | "\n", |
||
| 1433 | " AAK1_rnaseq ... AGAP2_cnv TSPAN31_cnv CDK4_cnv MARCH9_cnv \\\n", |
||
| 1434 | "0 1.5517 ... 0 0 0 0 \n", |
||
| 1435 | "1 0.5557 ... 0 0 0 0 \n", |
||
| 1436 | "2 0.5557 ... 0 0 0 0 \n", |
||
| 1437 | "3 0.6442 ... 2 2 2 2 \n", |
||
| 1438 | "4 0.6442 ... 2 2 2 2 \n", |
||
| 1439 | "... ... ... ... ... ... ... \n", |
||
| 1440 | "1037 -0.3841 ... 0 0 0 0 \n", |
||
| 1441 | "1038 -0.4479 ... -1 -1 -1 -1 \n", |
||
| 1442 | "1039 -0.2472 ... 0 0 0 0 \n", |
||
| 1443 | "1040 -0.5892 ... 0 0 0 0 \n", |
||
| 1444 | "1041 -0.1087 ... 0 0 0 0 \n", |
||
| 1445 | "\n", |
||
| 1446 | " CYP27B1_cnv METTL1_cnv TSFM_cnv AVIL_cnv CTDSP2_cnv RN7SKP65_cnv \n", |
||
| 1447 | "0 0 0 0 0 0 0 \n", |
||
| 1448 | "1 0 0 0 0 0 0 \n", |
||
| 1449 | "2 0 0 0 0 0 0 \n", |
||
| 1450 | "3 2 2 2 2 2 2 \n", |
||
| 1451 | "4 2 2 2 2 2 2 \n", |
||
| 1452 | "... ... ... ... ... ... ... \n", |
||
| 1453 | "1037 0 0 0 0 0 0 \n", |
||
| 1454 | "1038 -1 -1 -1 -1 -1 -1 \n", |
||
| 1455 | "1039 0 0 0 0 0 0 \n", |
||
| 1456 | "1040 0 0 0 0 0 0 \n", |
||
| 1457 | "1041 0 0 0 0 0 0 \n", |
||
| 1458 | "\n", |
||
| 1459 | "[1042 rows x 2891 columns]" |
||
| 1460 | ] |
||
| 1461 | }, |
||
| 1462 | "execution_count": 17, |
||
| 1463 | "metadata": {}, |
||
| 1464 | "output_type": "execute_result" |
||
| 1465 | } |
||
| 1466 | ], |
||
| 1467 | "source": [ |
||
| 1468 | "_" |
||
| 1469 | ] |
||
| 1470 | }, |
||
| 1471 | { |
||
| 1472 | "cell_type": "code", |
||
| 1473 | "execution_count": null, |
||
| 1474 | "metadata": {}, |
||
| 1475 | "outputs": [], |
||
| 1476 | "source": [] |
||
| 1477 | }, |
||
| 1478 | { |
||
| 1479 | "cell_type": "code", |
||
| 1480 | "execution_count": 30, |
||
| 1481 | "metadata": {}, |
||
| 1482 | "outputs": [], |
||
| 1483 | "source": [ |
||
| 1484 | "from scipy import stats\n", |
||
| 1485 | "\n", |
||
| 1486 | "slide_df = pd.read_csv(fname)\n", |
||
| 1487 | "rnaseq = slide_df[slide_df.columns[slide_df.columns.str.contains('_rnaseq')]]\n", |
||
| 1488 | "\n", |
||
| 1489 | "top_k=2000\n", |
||
| 1490 | "mad = stats.median_abs_deviation(rnaseq, axis=0)\n", |
||
| 1491 | "sort_idx = np.argsort(mad)[-top_k:]\n", |
||
| 1492 | "rnaseq = rnaseq[rnaseq.columns[sort_idx]]" |
||
| 1493 | ] |
||
| 1494 | }, |
||
| 1495 | { |
||
| 1496 | "cell_type": "code", |
||
| 1497 | "execution_count": 45, |
||
| 1498 | "metadata": {}, |
||
| 1499 | "outputs": [ |
||
| 1500 | { |
||
| 1501 | "data": { |
||
| 1502 | "text/html": [ |
||
| 1503 | "<div>\n", |
||
| 1504 | "<style scoped>\n", |
||
| 1505 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 1506 | " vertical-align: middle;\n", |
||
| 1507 | " }\n", |
||
| 1508 | "\n", |
||
| 1509 | " .dataframe tbody tr th {\n", |
||
| 1510 | " vertical-align: top;\n", |
||
| 1511 | " }\n", |
||
| 1512 | "\n", |
||
| 1513 | " .dataframe thead th {\n", |
||
| 1514 | " text-align: right;\n", |
||
| 1515 | " }\n", |
||
| 1516 | "</style>\n", |
||
| 1517 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 1518 | " <thead>\n", |
||
| 1519 | " <tr style=\"text-align: right;\">\n", |
||
| 1520 | " <th></th>\n", |
||
| 1521 | " <th>CLASRP_rnaseq</th>\n", |
||
| 1522 | " <th>NBAS_rnaseq</th>\n", |
||
| 1523 | " <th>ARL2BP_rnaseq</th>\n", |
||
| 1524 | " <th>TMEM199_rnaseq</th>\n", |
||
| 1525 | " <th>TTC37_rnaseq</th>\n", |
||
| 1526 | " <th>GTF2I_rnaseq</th>\n", |
||
| 1527 | " <th>STYX_rnaseq</th>\n", |
||
| 1528 | " <th>TSR3_rnaseq</th>\n", |
||
| 1529 | " <th>SEC61A1_rnaseq</th>\n", |
||
| 1530 | " <th>TRRAP_rnaseq</th>\n", |
||
| 1531 | " <th>...</th>\n", |
||
| 1532 | " <th>GET4_rnaseq</th>\n", |
||
| 1533 | " <th>BRD9_rnaseq</th>\n", |
||
| 1534 | " <th>NSUN2_rnaseq</th>\n", |
||
| 1535 | " <th>PYCRL_rnaseq</th>\n", |
||
| 1536 | " <th>HGH1_rnaseq</th>\n", |
||
| 1537 | " <th>PRUNE_rnaseq</th>\n", |
||
| 1538 | " <th>MAF1_rnaseq</th>\n", |
||
| 1539 | " <th>CCDC127_rnaseq</th>\n", |
||
| 1540 | " <th>EXOC3_rnaseq</th>\n", |
||
| 1541 | " <th>PUF60_rnaseq</th>\n", |
||
| 1542 | " </tr>\n", |
||
| 1543 | " </thead>\n", |
||
| 1544 | " <tbody>\n", |
||
| 1545 | " <tr>\n", |
||
| 1546 | " <th>0</th>\n", |
||
| 1547 | " <td>-0.5874</td>\n", |
||
| 1548 | " <td>0.8371</td>\n", |
||
| 1549 | " <td>0.7587</td>\n", |
||
| 1550 | " <td>0.2188</td>\n", |
||
| 1551 | " <td>-0.4040</td>\n", |
||
| 1552 | " <td>2.3916</td>\n", |
||
| 1553 | " <td>-0.7124</td>\n", |
||
| 1554 | " <td>-1.0035</td>\n", |
||
| 1555 | " <td>0.7356</td>\n", |
||
| 1556 | " <td>-0.0249</td>\n", |
||
| 1557 | " <td>...</td>\n", |
||
| 1558 | " <td>-0.1915</td>\n", |
||
| 1559 | " <td>0.3503</td>\n", |
||
| 1560 | " <td>-1.1848</td>\n", |
||
| 1561 | " <td>-1.4121</td>\n", |
||
| 1562 | " <td>-0.2389</td>\n", |
||
| 1563 | " <td>5.0110</td>\n", |
||
| 1564 | " <td>1.4287</td>\n", |
||
| 1565 | " <td>-0.3531</td>\n", |
||
| 1566 | " <td>-0.8503</td>\n", |
||
| 1567 | " <td>1.2995</td>\n", |
||
| 1568 | " </tr>\n", |
||
| 1569 | " <tr>\n", |
||
| 1570 | " <th>1</th>\n", |
||
| 1571 | " <td>0.2811</td>\n", |
||
| 1572 | " <td>0.2232</td>\n", |
||
| 1573 | " <td>0.9000</td>\n", |
||
| 1574 | " <td>3.2327</td>\n", |
||
| 1575 | " <td>-0.0096</td>\n", |
||
| 1576 | " <td>-0.4464</td>\n", |
||
| 1577 | " <td>0.5219</td>\n", |
||
| 1578 | " <td>0.3927</td>\n", |
||
| 1579 | " <td>-0.3513</td>\n", |
||
| 1580 | " <td>-0.7917</td>\n", |
||
| 1581 | " <td>...</td>\n", |
||
| 1582 | " <td>0.7627</td>\n", |
||
| 1583 | " <td>-0.6092</td>\n", |
||
| 1584 | " <td>0.1291</td>\n", |
||
| 1585 | " <td>1.7400</td>\n", |
||
| 1586 | " <td>-0.0250</td>\n", |
||
| 1587 | " <td>-0.1531</td>\n", |
||
| 1588 | " <td>0.5344</td>\n", |
||
| 1589 | " <td>-0.0012</td>\n", |
||
| 1590 | " <td>-0.9606</td>\n", |
||
| 1591 | " <td>2.0233</td>\n", |
||
| 1592 | " </tr>\n", |
||
| 1593 | " <tr>\n", |
||
| 1594 | " <th>2</th>\n", |
||
| 1595 | " <td>1.5665</td>\n", |
||
| 1596 | " <td>-0.4726</td>\n", |
||
| 1597 | " <td>0.1693</td>\n", |
||
| 1598 | " <td>0.9845</td>\n", |
||
| 1599 | " <td>-0.6740</td>\n", |
||
| 1600 | " <td>-0.3986</td>\n", |
||
| 1601 | " <td>-0.2289</td>\n", |
||
| 1602 | " <td>-0.2791</td>\n", |
||
| 1603 | " <td>-0.0646</td>\n", |
||
| 1604 | " <td>-0.4431</td>\n", |
||
| 1605 | " <td>...</td>\n", |
||
| 1606 | " <td>4.4123</td>\n", |
||
| 1607 | " <td>1.4417</td>\n", |
||
| 1608 | " <td>-0.5196</td>\n", |
||
| 1609 | " <td>-1.3030</td>\n", |
||
| 1610 | " <td>-1.1373</td>\n", |
||
| 1611 | " <td>4.6041</td>\n", |
||
| 1612 | " <td>-1.0135</td>\n", |
||
| 1613 | " <td>1.3589</td>\n", |
||
| 1614 | " <td>2.6994</td>\n", |
||
| 1615 | " <td>-0.3068</td>\n", |
||
| 1616 | " </tr>\n", |
||
| 1617 | " <tr>\n", |
||
| 1618 | " <th>3</th>\n", |
||
| 1619 | " <td>0.6169</td>\n", |
||
| 1620 | " <td>-0.3266</td>\n", |
||
| 1621 | " <td>-0.3082</td>\n", |
||
| 1622 | " <td>-0.2220</td>\n", |
||
| 1623 | " <td>0.5305</td>\n", |
||
| 1624 | " <td>0.5360</td>\n", |
||
| 1625 | " <td>0.2785</td>\n", |
||
| 1626 | " <td>-1.0873</td>\n", |
||
| 1627 | " <td>-1.0712</td>\n", |
||
| 1628 | " <td>-0.3184</td>\n", |
||
| 1629 | " <td>...</td>\n", |
||
| 1630 | " <td>0.7665</td>\n", |
||
| 1631 | " <td>-0.3344</td>\n", |
||
| 1632 | " <td>0.0695</td>\n", |
||
| 1633 | " <td>0.0040</td>\n", |
||
| 1634 | " <td>0.2291</td>\n", |
||
| 1635 | " <td>3.6034</td>\n", |
||
| 1636 | " <td>0.1774</td>\n", |
||
| 1637 | " <td>-0.2766</td>\n", |
||
| 1638 | " <td>0.5080</td>\n", |
||
| 1639 | " <td>0.6178</td>\n", |
||
| 1640 | " </tr>\n", |
||
| 1641 | " <tr>\n", |
||
| 1642 | " <th>4</th>\n", |
||
| 1643 | " <td>0.6406</td>\n", |
||
| 1644 | " <td>-1.0330</td>\n", |
||
| 1645 | " <td>-0.6522</td>\n", |
||
| 1646 | " <td>0.1727</td>\n", |
||
| 1647 | " <td>-0.7455</td>\n", |
||
| 1648 | " <td>-0.6040</td>\n", |
||
| 1649 | " <td>0.2553</td>\n", |
||
| 1650 | " <td>1.0504</td>\n", |
||
| 1651 | " <td>1.0583</td>\n", |
||
| 1652 | " <td>-0.2884</td>\n", |
||
| 1653 | " <td>...</td>\n", |
||
| 1654 | " <td>3.3807</td>\n", |
||
| 1655 | " <td>0.3364</td>\n", |
||
| 1656 | " <td>-0.2792</td>\n", |
||
| 1657 | " <td>4.8566</td>\n", |
||
| 1658 | " <td>7.9296</td>\n", |
||
| 1659 | " <td>1.6951</td>\n", |
||
| 1660 | " <td>5.8943</td>\n", |
||
| 1661 | " <td>1.3652</td>\n", |
||
| 1662 | " <td>-0.8062</td>\n", |
||
| 1663 | " <td>9.2417</td>\n", |
||
| 1664 | " </tr>\n", |
||
| 1665 | " <tr>\n", |
||
| 1666 | " <th>...</th>\n", |
||
| 1667 | " <td>...</td>\n", |
||
| 1668 | " <td>...</td>\n", |
||
| 1669 | " <td>...</td>\n", |
||
| 1670 | " <td>...</td>\n", |
||
| 1671 | " <td>...</td>\n", |
||
| 1672 | " <td>...</td>\n", |
||
| 1673 | " <td>...</td>\n", |
||
| 1674 | " <td>...</td>\n", |
||
| 1675 | " <td>...</td>\n", |
||
| 1676 | " <td>...</td>\n", |
||
| 1677 | " <td>...</td>\n", |
||
| 1678 | " <td>...</td>\n", |
||
| 1679 | " <td>...</td>\n", |
||
| 1680 | " <td>...</td>\n", |
||
| 1681 | " <td>...</td>\n", |
||
| 1682 | " <td>...</td>\n", |
||
| 1683 | " <td>...</td>\n", |
||
| 1684 | " <td>...</td>\n", |
||
| 1685 | " <td>...</td>\n", |
||
| 1686 | " <td>...</td>\n", |
||
| 1687 | " <td>...</td>\n", |
||
| 1688 | " </tr>\n", |
||
| 1689 | " <tr>\n", |
||
| 1690 | " <th>511</th>\n", |
||
| 1691 | " <td>0.5640</td>\n", |
||
| 1692 | " <td>0.1255</td>\n", |
||
| 1693 | " <td>-1.3364</td>\n", |
||
| 1694 | " <td>-0.8430</td>\n", |
||
| 1695 | " <td>0.4406</td>\n", |
||
| 1696 | " <td>-0.9735</td>\n", |
||
| 1697 | " <td>-1.4547</td>\n", |
||
| 1698 | " <td>-0.1983</td>\n", |
||
| 1699 | " <td>-0.5259</td>\n", |
||
| 1700 | " <td>-0.2029</td>\n", |
||
| 1701 | " <td>...</td>\n", |
||
| 1702 | " <td>3.0311</td>\n", |
||
| 1703 | " <td>3.4963</td>\n", |
||
| 1704 | " <td>2.5079</td>\n", |
||
| 1705 | " <td>0.0556</td>\n", |
||
| 1706 | " <td>0.5691</td>\n", |
||
| 1707 | " <td>0.1104</td>\n", |
||
| 1708 | " <td>-0.3776</td>\n", |
||
| 1709 | " <td>2.6136</td>\n", |
||
| 1710 | " <td>3.4259</td>\n", |
||
| 1711 | " <td>-0.8442</td>\n", |
||
| 1712 | " </tr>\n", |
||
| 1713 | " <tr>\n", |
||
| 1714 | " <th>512</th>\n", |
||
| 1715 | " <td>1.2336</td>\n", |
||
| 1716 | " <td>0.1902</td>\n", |
||
| 1717 | " <td>-1.3500</td>\n", |
||
| 1718 | " <td>-0.3472</td>\n", |
||
| 1719 | " <td>0.4549</td>\n", |
||
| 1720 | " <td>-0.6806</td>\n", |
||
| 1721 | " <td>-1.1291</td>\n", |
||
| 1722 | " <td>1.0677</td>\n", |
||
| 1723 | " <td>1.1586</td>\n", |
||
| 1724 | " <td>0.1959</td>\n", |
||
| 1725 | " <td>...</td>\n", |
||
| 1726 | " <td>0.5573</td>\n", |
||
| 1727 | " <td>-0.7546</td>\n", |
||
| 1728 | " <td>0.8104</td>\n", |
||
| 1729 | " <td>0.1239</td>\n", |
||
| 1730 | " <td>0.0985</td>\n", |
||
| 1731 | " <td>2.9026</td>\n", |
||
| 1732 | " <td>0.0173</td>\n", |
||
| 1733 | " <td>0.3492</td>\n", |
||
| 1734 | " <td>2.5703</td>\n", |
||
| 1735 | " <td>1.0690</td>\n", |
||
| 1736 | " </tr>\n", |
||
| 1737 | " <tr>\n", |
||
| 1738 | " <th>513</th>\n", |
||
| 1739 | " <td>1.8148</td>\n", |
||
| 1740 | " <td>-0.8502</td>\n", |
||
| 1741 | " <td>-0.0628</td>\n", |
||
| 1742 | " <td>-0.7776</td>\n", |
||
| 1743 | " <td>0.6452</td>\n", |
||
| 1744 | " <td>-0.4622</td>\n", |
||
| 1745 | " <td>-1.2732</td>\n", |
||
| 1746 | " <td>1.8145</td>\n", |
||
| 1747 | " <td>-0.8767</td>\n", |
||
| 1748 | " <td>-0.2980</td>\n", |
||
| 1749 | " <td>...</td>\n", |
||
| 1750 | " <td>1.4671</td>\n", |
||
| 1751 | " <td>2.2343</td>\n", |
||
| 1752 | " <td>2.1466</td>\n", |
||
| 1753 | " <td>0.7868</td>\n", |
||
| 1754 | " <td>0.6893</td>\n", |
||
| 1755 | " <td>0.1571</td>\n", |
||
| 1756 | " <td>0.9686</td>\n", |
||
| 1757 | " <td>3.0870</td>\n", |
||
| 1758 | " <td>5.6169</td>\n", |
||
| 1759 | " <td>0.5300</td>\n", |
||
| 1760 | " </tr>\n", |
||
| 1761 | " <tr>\n", |
||
| 1762 | " <th>514</th>\n", |
||
| 1763 | " <td>0.0569</td>\n", |
||
| 1764 | " <td>-0.4511</td>\n", |
||
| 1765 | " <td>3.8784</td>\n", |
||
| 1766 | " <td>0.2609</td>\n", |
||
| 1767 | " <td>0.9393</td>\n", |
||
| 1768 | " <td>0.5776</td>\n", |
||
| 1769 | " <td>-0.9469</td>\n", |
||
| 1770 | " <td>2.9500</td>\n", |
||
| 1771 | " <td>-0.9261</td>\n", |
||
| 1772 | " <td>2.4218</td>\n", |
||
| 1773 | " <td>...</td>\n", |
||
| 1774 | " <td>5.0440</td>\n", |
||
| 1775 | " <td>0.0862</td>\n", |
||
| 1776 | " <td>0.1431</td>\n", |
||
| 1777 | " <td>-0.7761</td>\n", |
||
| 1778 | " <td>-0.8430</td>\n", |
||
| 1779 | " <td>0.2311</td>\n", |
||
| 1780 | " <td>-0.5913</td>\n", |
||
| 1781 | " <td>1.4958</td>\n", |
||
| 1782 | " <td>2.1736</td>\n", |
||
| 1783 | " <td>-0.5699</td>\n", |
||
| 1784 | " </tr>\n", |
||
| 1785 | " <tr>\n", |
||
| 1786 | " <th>515</th>\n", |
||
| 1787 | " <td>1.9203</td>\n", |
||
| 1788 | " <td>-0.0634</td>\n", |
||
| 1789 | " <td>-0.7142</td>\n", |
||
| 1790 | " <td>-1.3296</td>\n", |
||
| 1791 | " <td>0.3966</td>\n", |
||
| 1792 | " <td>1.0089</td>\n", |
||
| 1793 | " <td>-0.7931</td>\n", |
||
| 1794 | " <td>0.8513</td>\n", |
||
| 1795 | " <td>0.7651</td>\n", |
||
| 1796 | " <td>0.2217</td>\n", |
||
| 1797 | " <td>...</td>\n", |
||
| 1798 | " <td>2.2952</td>\n", |
||
| 1799 | " <td>0.3985</td>\n", |
||
| 1800 | " <td>-0.0072</td>\n", |
||
| 1801 | " <td>0.0683</td>\n", |
||
| 1802 | " <td>-0.8047</td>\n", |
||
| 1803 | " <td>-0.2712</td>\n", |
||
| 1804 | " <td>-0.5864</td>\n", |
||
| 1805 | " <td>-0.2393</td>\n", |
||
| 1806 | " <td>1.8585</td>\n", |
||
| 1807 | " <td>-1.0489</td>\n", |
||
| 1808 | " </tr>\n", |
||
| 1809 | " </tbody>\n", |
||
| 1810 | "</table>\n", |
||
| 1811 | "<p>516 rows × 2000 columns</p>\n", |
||
| 1812 | "</div>" |
||
| 1813 | ], |
||
| 1814 | "text/plain": [ |
||
| 1815 | " CLASRP_rnaseq NBAS_rnaseq ARL2BP_rnaseq TMEM199_rnaseq TTC37_rnaseq \\\n", |
||
| 1816 | "0 -0.5874 0.8371 0.7587 0.2188 -0.4040 \n", |
||
| 1817 | "1 0.2811 0.2232 0.9000 3.2327 -0.0096 \n", |
||
| 1818 | "2 1.5665 -0.4726 0.1693 0.9845 -0.6740 \n", |
||
| 1819 | "3 0.6169 -0.3266 -0.3082 -0.2220 0.5305 \n", |
||
| 1820 | "4 0.6406 -1.0330 -0.6522 0.1727 -0.7455 \n", |
||
| 1821 | ".. ... ... ... ... ... \n", |
||
| 1822 | "511 0.5640 0.1255 -1.3364 -0.8430 0.4406 \n", |
||
| 1823 | "512 1.2336 0.1902 -1.3500 -0.3472 0.4549 \n", |
||
| 1824 | "513 1.8148 -0.8502 -0.0628 -0.7776 0.6452 \n", |
||
| 1825 | "514 0.0569 -0.4511 3.8784 0.2609 0.9393 \n", |
||
| 1826 | "515 1.9203 -0.0634 -0.7142 -1.3296 0.3966 \n", |
||
| 1827 | "\n", |
||
| 1828 | " GTF2I_rnaseq STYX_rnaseq TSR3_rnaseq SEC61A1_rnaseq TRRAP_rnaseq \\\n", |
||
| 1829 | "0 2.3916 -0.7124 -1.0035 0.7356 -0.0249 \n", |
||
| 1830 | "1 -0.4464 0.5219 0.3927 -0.3513 -0.7917 \n", |
||
| 1831 | "2 -0.3986 -0.2289 -0.2791 -0.0646 -0.4431 \n", |
||
| 1832 | "3 0.5360 0.2785 -1.0873 -1.0712 -0.3184 \n", |
||
| 1833 | "4 -0.6040 0.2553 1.0504 1.0583 -0.2884 \n", |
||
| 1834 | ".. ... ... ... ... ... \n", |
||
| 1835 | "511 -0.9735 -1.4547 -0.1983 -0.5259 -0.2029 \n", |
||
| 1836 | "512 -0.6806 -1.1291 1.0677 1.1586 0.1959 \n", |
||
| 1837 | "513 -0.4622 -1.2732 1.8145 -0.8767 -0.2980 \n", |
||
| 1838 | "514 0.5776 -0.9469 2.9500 -0.9261 2.4218 \n", |
||
| 1839 | "515 1.0089 -0.7931 0.8513 0.7651 0.2217 \n", |
||
| 1840 | "\n", |
||
| 1841 | " ... GET4_rnaseq BRD9_rnaseq NSUN2_rnaseq PYCRL_rnaseq HGH1_rnaseq \\\n", |
||
| 1842 | "0 ... -0.1915 0.3503 -1.1848 -1.4121 -0.2389 \n", |
||
| 1843 | "1 ... 0.7627 -0.6092 0.1291 1.7400 -0.0250 \n", |
||
| 1844 | "2 ... 4.4123 1.4417 -0.5196 -1.3030 -1.1373 \n", |
||
| 1845 | "3 ... 0.7665 -0.3344 0.0695 0.0040 0.2291 \n", |
||
| 1846 | "4 ... 3.3807 0.3364 -0.2792 4.8566 7.9296 \n", |
||
| 1847 | ".. ... ... ... ... ... ... \n", |
||
| 1848 | "511 ... 3.0311 3.4963 2.5079 0.0556 0.5691 \n", |
||
| 1849 | "512 ... 0.5573 -0.7546 0.8104 0.1239 0.0985 \n", |
||
| 1850 | "513 ... 1.4671 2.2343 2.1466 0.7868 0.6893 \n", |
||
| 1851 | "514 ... 5.0440 0.0862 0.1431 -0.7761 -0.8430 \n", |
||
| 1852 | "515 ... 2.2952 0.3985 -0.0072 0.0683 -0.8047 \n", |
||
| 1853 | "\n", |
||
| 1854 | " PRUNE_rnaseq MAF1_rnaseq CCDC127_rnaseq EXOC3_rnaseq PUF60_rnaseq \n", |
||
| 1855 | "0 5.0110 1.4287 -0.3531 -0.8503 1.2995 \n", |
||
| 1856 | "1 -0.1531 0.5344 -0.0012 -0.9606 2.0233 \n", |
||
| 1857 | "2 4.6041 -1.0135 1.3589 2.6994 -0.3068 \n", |
||
| 1858 | "3 3.6034 0.1774 -0.2766 0.5080 0.6178 \n", |
||
| 1859 | "4 1.6951 5.8943 1.3652 -0.8062 9.2417 \n", |
||
| 1860 | ".. ... ... ... ... ... \n", |
||
| 1861 | "511 0.1104 -0.3776 2.6136 3.4259 -0.8442 \n", |
||
| 1862 | "512 2.9026 0.0173 0.3492 2.5703 1.0690 \n", |
||
| 1863 | "513 0.1571 0.9686 3.0870 5.6169 0.5300 \n", |
||
| 1864 | "514 0.2311 -0.5913 1.4958 2.1736 -0.5699 \n", |
||
| 1865 | "515 -0.2712 -0.5864 -0.2393 1.8585 -1.0489 \n", |
||
| 1866 | "\n", |
||
| 1867 | "[516 rows x 2000 columns]" |
||
| 1868 | ] |
||
| 1869 | }, |
||
| 1870 | "execution_count": 45, |
||
| 1871 | "metadata": {}, |
||
| 1872 | "output_type": "execute_result" |
||
| 1873 | } |
||
| 1874 | ], |
||
| 1875 | "source": [ |
||
| 1876 | "rnaseq" |
||
| 1877 | ] |
||
| 1878 | }, |
||
| 1879 | { |
||
| 1880 | "cell_type": "code", |
||
| 1881 | "execution_count": 51, |
||
| 1882 | "metadata": {}, |
||
| 1883 | "outputs": [ |
||
| 1884 | { |
||
| 1885 | "ename": "ModuleNotFoundError", |
||
| 1886 | "evalue": "No module named 'torch'", |
||
| 1887 | "output_type": "error", |
||
| 1888 | "traceback": [ |
||
| 1889 | "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", |
||
| 1890 | "\u001b[0;31mModuleNotFoundError\u001b[0m Traceback (most recent call last)", |
||
| 1891 | "\u001b[0;32m<ipython-input-51-eb42ca6e4af3>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0;32mimport\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m", |
||
| 1892 | "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'torch'" |
||
| 1893 | ] |
||
| 1894 | } |
||
| 1895 | ], |
||
| 1896 | "source": [ |
||
| 1897 | "import torch" |
||
| 1898 | ] |
||
| 1899 | }, |
||
| 1900 | { |
||
| 1901 | "cell_type": "code", |
||
| 1902 | "execution_count": null, |
||
| 1903 | "metadata": {}, |
||
| 1904 | "outputs": [], |
||
| 1905 | "source": [] |
||
| 1906 | }, |
||
| 1907 | { |
||
| 1908 | "cell_type": "code", |
||
| 1909 | "execution_count": 20, |
||
| 1910 | "metadata": {}, |
||
| 1911 | "outputs": [ |
||
| 1912 | { |
||
| 1913 | "data": { |
||
| 1914 | "text/plain": [ |
||
| 1915 | "Index(['UBE2Q2P2_rnaseq', 'SSX9_rnaseq', 'CXORF67_rnaseq', 'EFCAB8_rnaseq',\n", |
||
| 1916 | " 'SDR16C6P_rnaseq', 'EFCAB12_rnaseq', 'A1BG_rnaseq', 'A1CF_rnaseq',\n", |
||
| 1917 | " 'RBFOX1_rnaseq', 'GGACT_rnaseq',\n", |
||
| 1918 | " ...\n", |
||
| 1919 | " 'ZWINT_rnaseq', 'ZXDA_rnaseq', 'ZXDB_rnaseq', 'ZXDC_rnaseq',\n", |
||
| 1920 | " 'ZYG11A_rnaseq', 'ZYG11B_rnaseq', 'ZYX_rnaseq', 'ZZEF1_rnaseq',\n", |
||
| 1921 | " 'ZZZ3_rnaseq', 'TPTEP1_rnaseq'],\n", |
||
| 1922 | " dtype='object', length=18345)" |
||
| 1923 | ] |
||
| 1924 | }, |
||
| 1925 | "execution_count": 20, |
||
| 1926 | "metadata": {}, |
||
| 1927 | "output_type": "execute_result" |
||
| 1928 | } |
||
| 1929 | ], |
||
| 1930 | "source": [ |
||
| 1931 | "slide_df[slide_df.columns.str.contains('_rnaseq')]" |
||
| 1932 | ] |
||
| 1933 | }, |
||
| 1934 | { |
||
| 1935 | "cell_type": "code", |
||
| 1936 | "execution_count": 24, |
||
| 1937 | "metadata": {}, |
||
| 1938 | "outputs": [], |
||
| 1939 | "source": [ |
||
| 1940 | "slide_df = pd.read_csv(fname)\n", |
||
| 1941 | "slide_df = slide_df[slide_df.columns[slide_df.columns.str.contains('_rnaseq')]]" |
||
| 1942 | ] |
||
| 1943 | }, |
||
| 1944 | { |
||
| 1945 | "cell_type": "code", |
||
| 1946 | "execution_count": 25, |
||
| 1947 | "metadata": {}, |
||
| 1948 | "outputs": [ |
||
| 1949 | { |
||
| 1950 | "data": { |
||
| 1951 | "text/html": [ |
||
| 1952 | "<div>\n", |
||
| 1953 | "<style scoped>\n", |
||
| 1954 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 1955 | " vertical-align: middle;\n", |
||
| 1956 | " }\n", |
||
| 1957 | "\n", |
||
| 1958 | " .dataframe tbody tr th {\n", |
||
| 1959 | " vertical-align: top;\n", |
||
| 1960 | " }\n", |
||
| 1961 | "\n", |
||
| 1962 | " .dataframe thead th {\n", |
||
| 1963 | " text-align: right;\n", |
||
| 1964 | " }\n", |
||
| 1965 | "</style>\n", |
||
| 1966 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 1967 | " <thead>\n", |
||
| 1968 | " <tr style=\"text-align: right;\">\n", |
||
| 1969 | " <th></th>\n", |
||
| 1970 | " <th>UBE2Q2P2_rnaseq</th>\n", |
||
| 1971 | " <th>SSX9_rnaseq</th>\n", |
||
| 1972 | " <th>CXORF67_rnaseq</th>\n", |
||
| 1973 | " <th>EFCAB8_rnaseq</th>\n", |
||
| 1974 | " <th>SDR16C6P_rnaseq</th>\n", |
||
| 1975 | " <th>EFCAB12_rnaseq</th>\n", |
||
| 1976 | " <th>A1BG_rnaseq</th>\n", |
||
| 1977 | " <th>A1CF_rnaseq</th>\n", |
||
| 1978 | " <th>RBFOX1_rnaseq</th>\n", |
||
| 1979 | " <th>GGACT_rnaseq</th>\n", |
||
| 1980 | " <th>...</th>\n", |
||
| 1981 | " <th>ZWINT_rnaseq</th>\n", |
||
| 1982 | " <th>ZXDA_rnaseq</th>\n", |
||
| 1983 | " <th>ZXDB_rnaseq</th>\n", |
||
| 1984 | " <th>ZXDC_rnaseq</th>\n", |
||
| 1985 | " <th>ZYG11A_rnaseq</th>\n", |
||
| 1986 | " <th>ZYG11B_rnaseq</th>\n", |
||
| 1987 | " <th>ZYX_rnaseq</th>\n", |
||
| 1988 | " <th>ZZEF1_rnaseq</th>\n", |
||
| 1989 | " <th>ZZZ3_rnaseq</th>\n", |
||
| 1990 | " <th>TPTEP1_rnaseq</th>\n", |
||
| 1991 | " </tr>\n", |
||
| 1992 | " </thead>\n", |
||
| 1993 | " <tbody>\n", |
||
| 1994 | " <tr>\n", |
||
| 1995 | " <th>0</th>\n", |
||
| 1996 | " <td>-0.3291</td>\n", |
||
| 1997 | " <td>-0.1379</td>\n", |
||
| 1998 | " <td>-0.1805</td>\n", |
||
| 1999 | " <td>-0.0869</td>\n", |
||
| 2000 | " <td>-0.3317</td>\n", |
||
| 2001 | " <td>-0.1661</td>\n", |
||
| 2002 | " <td>-0.1483</td>\n", |
||
| 2003 | " <td>-0.1371</td>\n", |
||
| 2004 | " <td>-0.2260</td>\n", |
||
| 2005 | " <td>-0.5346</td>\n", |
||
| 2006 | " <td>...</td>\n", |
||
| 2007 | " <td>-0.7082</td>\n", |
||
| 2008 | " <td>0.6149</td>\n", |
||
| 2009 | " <td>0.5725</td>\n", |
||
| 2010 | " <td>0.2889</td>\n", |
||
| 2011 | " <td>-0.5255</td>\n", |
||
| 2012 | " <td>-0.2205</td>\n", |
||
| 2013 | " <td>-0.7847</td>\n", |
||
| 2014 | " <td>-0.2296</td>\n", |
||
| 2015 | " <td>-0.0897</td>\n", |
||
| 2016 | " <td>0.1457</td>\n", |
||
| 2017 | " </tr>\n", |
||
| 2018 | " <tr>\n", |
||
| 2019 | " <th>1</th>\n", |
||
| 2020 | " <td>-0.8531</td>\n", |
||
| 2021 | " <td>-0.1379</td>\n", |
||
| 2022 | " <td>-0.1805</td>\n", |
||
| 2023 | " <td>-0.2629</td>\n", |
||
| 2024 | " <td>-0.3317</td>\n", |
||
| 2025 | " <td>-0.2317</td>\n", |
||
| 2026 | " <td>-0.5528</td>\n", |
||
| 2027 | " <td>-0.1476</td>\n", |
||
| 2028 | " <td>-0.2508</td>\n", |
||
| 2029 | " <td>0.6921</td>\n", |
||
| 2030 | " <td>...</td>\n", |
||
| 2031 | " <td>0.9291</td>\n", |
||
| 2032 | " <td>0.5118</td>\n", |
||
| 2033 | " <td>-0.1673</td>\n", |
||
| 2034 | " <td>-0.8006</td>\n", |
||
| 2035 | " <td>-0.4348</td>\n", |
||
| 2036 | " <td>-1.7113</td>\n", |
||
| 2037 | " <td>0.7466</td>\n", |
||
| 2038 | " <td>-0.1563</td>\n", |
||
| 2039 | " <td>-0.9102</td>\n", |
||
| 2040 | " <td>-0.5005</td>\n", |
||
| 2041 | " </tr>\n", |
||
| 2042 | " <tr>\n", |
||
| 2043 | " <th>2</th>\n", |
||
| 2044 | " <td>-0.7262</td>\n", |
||
| 2045 | " <td>0.3883</td>\n", |
||
| 2046 | " <td>0.4908</td>\n", |
||
| 2047 | " <td>-0.0666</td>\n", |
||
| 2048 | " <td>-0.3317</td>\n", |
||
| 2049 | " <td>-0.3948</td>\n", |
||
| 2050 | " <td>0.0021</td>\n", |
||
| 2051 | " <td>-0.1476</td>\n", |
||
| 2052 | " <td>-0.2508</td>\n", |
||
| 2053 | " <td>-0.0800</td>\n", |
||
| 2054 | " <td>...</td>\n", |
||
| 2055 | " <td>0.2957</td>\n", |
||
| 2056 | " <td>-0.4399</td>\n", |
||
| 2057 | " <td>-0.2751</td>\n", |
||
| 2058 | " <td>-0.4668</td>\n", |
||
| 2059 | " <td>0.1222</td>\n", |
||
| 2060 | " <td>0.3555</td>\n", |
||
| 2061 | " <td>1.4078</td>\n", |
||
| 2062 | " <td>-0.1592</td>\n", |
||
| 2063 | " <td>-0.2276</td>\n", |
||
| 2064 | " <td>-0.3931</td>\n", |
||
| 2065 | " </tr>\n", |
||
| 2066 | " <tr>\n", |
||
| 2067 | " <th>3</th>\n", |
||
| 2068 | " <td>-1.0590</td>\n", |
||
| 2069 | " <td>-0.1379</td>\n", |
||
| 2070 | " <td>-0.1805</td>\n", |
||
| 2071 | " <td>-0.0959</td>\n", |
||
| 2072 | " <td>-0.3317</td>\n", |
||
| 2073 | " <td>-0.3372</td>\n", |
||
| 2074 | " <td>-0.1061</td>\n", |
||
| 2075 | " <td>-0.1476</td>\n", |
||
| 2076 | " <td>-0.2508</td>\n", |
||
| 2077 | " <td>-0.5641</td>\n", |
||
| 2078 | " <td>...</td>\n", |
||
| 2079 | " <td>-0.9962</td>\n", |
||
| 2080 | " <td>1.4844</td>\n", |
||
| 2081 | " <td>0.9748</td>\n", |
||
| 2082 | " <td>0.7481</td>\n", |
||
| 2083 | " <td>-0.7049</td>\n", |
||
| 2084 | " <td>-0.2617</td>\n", |
||
| 2085 | " <td>-0.2934</td>\n", |
||
| 2086 | " <td>1.1243</td>\n", |
||
| 2087 | " <td>0.0823</td>\n", |
||
| 2088 | " <td>0.8831</td>\n", |
||
| 2089 | " </tr>\n", |
||
| 2090 | " <tr>\n", |
||
| 2091 | " <th>4</th>\n", |
||
| 2092 | " <td>-0.7257</td>\n", |
||
| 2093 | " <td>-0.1379</td>\n", |
||
| 2094 | " <td>-0.1805</td>\n", |
||
| 2095 | " <td>-0.1756</td>\n", |
||
| 2096 | " <td>-0.3317</td>\n", |
||
| 2097 | " <td>-0.3778</td>\n", |
||
| 2098 | " <td>0.1119</td>\n", |
||
| 2099 | " <td>-0.1476</td>\n", |
||
| 2100 | " <td>1.2515</td>\n", |
||
| 2101 | " <td>-1.0113</td>\n", |
||
| 2102 | " <td>...</td>\n", |
||
| 2103 | " <td>1.7870</td>\n", |
||
| 2104 | " <td>-0.0462</td>\n", |
||
| 2105 | " <td>1.8418</td>\n", |
||
| 2106 | " <td>-0.9922</td>\n", |
||
| 2107 | " <td>-0.7090</td>\n", |
||
| 2108 | " <td>-1.0285</td>\n", |
||
| 2109 | " <td>0.6567</td>\n", |
||
| 2110 | " <td>-1.0377</td>\n", |
||
| 2111 | " <td>-1.1277</td>\n", |
||
| 2112 | " <td>-0.5026</td>\n", |
||
| 2113 | " </tr>\n", |
||
| 2114 | " <tr>\n", |
||
| 2115 | " <th>...</th>\n", |
||
| 2116 | " <td>...</td>\n", |
||
| 2117 | " <td>...</td>\n", |
||
| 2118 | " <td>...</td>\n", |
||
| 2119 | " <td>...</td>\n", |
||
| 2120 | " <td>...</td>\n", |
||
| 2121 | " <td>...</td>\n", |
||
| 2122 | " <td>...</td>\n", |
||
| 2123 | " <td>...</td>\n", |
||
| 2124 | " <td>...</td>\n", |
||
| 2125 | " <td>...</td>\n", |
||
| 2126 | " <td>...</td>\n", |
||
| 2127 | " <td>...</td>\n", |
||
| 2128 | " <td>...</td>\n", |
||
| 2129 | " <td>...</td>\n", |
||
| 2130 | " <td>...</td>\n", |
||
| 2131 | " <td>...</td>\n", |
||
| 2132 | " <td>...</td>\n", |
||
| 2133 | " <td>...</td>\n", |
||
| 2134 | " <td>...</td>\n", |
||
| 2135 | " <td>...</td>\n", |
||
| 2136 | " <td>...</td>\n", |
||
| 2137 | " </tr>\n", |
||
| 2138 | " <tr>\n", |
||
| 2139 | " <th>511</th>\n", |
||
| 2140 | " <td>0.5308</td>\n", |
||
| 2141 | " <td>-0.1379</td>\n", |
||
| 2142 | " <td>-0.1805</td>\n", |
||
| 2143 | " <td>-0.2629</td>\n", |
||
| 2144 | " <td>-0.3317</td>\n", |
||
| 2145 | " <td>-0.2827</td>\n", |
||
| 2146 | " <td>-0.6045</td>\n", |
||
| 2147 | " <td>-0.1476</td>\n", |
||
| 2148 | " <td>-0.2508</td>\n", |
||
| 2149 | " <td>-0.2014</td>\n", |
||
| 2150 | " <td>...</td>\n", |
||
| 2151 | " <td>-0.5331</td>\n", |
||
| 2152 | " <td>-0.4205</td>\n", |
||
| 2153 | " <td>-0.3773</td>\n", |
||
| 2154 | " <td>0.0551</td>\n", |
||
| 2155 | " <td>-0.5660</td>\n", |
||
| 2156 | " <td>-0.5123</td>\n", |
||
| 2157 | " <td>0.1254</td>\n", |
||
| 2158 | " <td>0.2124</td>\n", |
||
| 2159 | " <td>-0.6375</td>\n", |
||
| 2160 | " <td>1.4712</td>\n", |
||
| 2161 | " </tr>\n", |
||
| 2162 | " <tr>\n", |
||
| 2163 | " <th>512</th>\n", |
||
| 2164 | " <td>-0.5021</td>\n", |
||
| 2165 | " <td>-0.1379</td>\n", |
||
| 2166 | " <td>-0.0120</td>\n", |
||
| 2167 | " <td>1.7408</td>\n", |
||
| 2168 | " <td>-0.3317</td>\n", |
||
| 2169 | " <td>-0.2152</td>\n", |
||
| 2170 | " <td>0.7495</td>\n", |
||
| 2171 | " <td>1.8708</td>\n", |
||
| 2172 | " <td>-0.1178</td>\n", |
||
| 2173 | " <td>-1.3502</td>\n", |
||
| 2174 | " <td>...</td>\n", |
||
| 2175 | " <td>-0.3624</td>\n", |
||
| 2176 | " <td>0.0588</td>\n", |
||
| 2177 | " <td>-0.1157</td>\n", |
||
| 2178 | " <td>1.2831</td>\n", |
||
| 2179 | " <td>-0.0555</td>\n", |
||
| 2180 | " <td>-0.3620</td>\n", |
||
| 2181 | " <td>-0.4242</td>\n", |
||
| 2182 | " <td>1.6937</td>\n", |
||
| 2183 | " <td>-0.4990</td>\n", |
||
| 2184 | " <td>2.2944</td>\n", |
||
| 2185 | " </tr>\n", |
||
| 2186 | " <tr>\n", |
||
| 2187 | " <th>513</th>\n", |
||
| 2188 | " <td>5.2714</td>\n", |
||
| 2189 | " <td>-0.1379</td>\n", |
||
| 2190 | " <td>-0.1805</td>\n", |
||
| 2191 | " <td>0.1753</td>\n", |
||
| 2192 | " <td>-0.3317</td>\n", |
||
| 2193 | " <td>-0.2325</td>\n", |
||
| 2194 | " <td>0.5863</td>\n", |
||
| 2195 | " <td>-0.1476</td>\n", |
||
| 2196 | " <td>-0.0185</td>\n", |
||
| 2197 | " <td>-0.3172</td>\n", |
||
| 2198 | " <td>...</td>\n", |
||
| 2199 | " <td>-0.9598</td>\n", |
||
| 2200 | " <td>-1.5823</td>\n", |
||
| 2201 | " <td>-1.3015</td>\n", |
||
| 2202 | " <td>0.4371</td>\n", |
||
| 2203 | " <td>-0.6739</td>\n", |
||
| 2204 | " <td>-1.4417</td>\n", |
||
| 2205 | " <td>-0.9613</td>\n", |
||
| 2206 | " <td>0.4167</td>\n", |
||
| 2207 | " <td>-1.4631</td>\n", |
||
| 2208 | " <td>-0.5035</td>\n", |
||
| 2209 | " </tr>\n", |
||
| 2210 | " <tr>\n", |
||
| 2211 | " <th>514</th>\n", |
||
| 2212 | " <td>0.6290</td>\n", |
||
| 2213 | " <td>-0.1379</td>\n", |
||
| 2214 | " <td>0.1131</td>\n", |
||
| 2215 | " <td>-0.0667</td>\n", |
||
| 2216 | " <td>1.5316</td>\n", |
||
| 2217 | " <td>-0.3634</td>\n", |
||
| 2218 | " <td>0.3730</td>\n", |
||
| 2219 | " <td>-0.1476</td>\n", |
||
| 2220 | " <td>-0.2361</td>\n", |
||
| 2221 | " <td>-1.7106</td>\n", |
||
| 2222 | " <td>...</td>\n", |
||
| 2223 | " <td>-0.5337</td>\n", |
||
| 2224 | " <td>4.2234</td>\n", |
||
| 2225 | " <td>0.9716</td>\n", |
||
| 2226 | " <td>0.6699</td>\n", |
||
| 2227 | " <td>-0.8134</td>\n", |
||
| 2228 | " <td>-0.2453</td>\n", |
||
| 2229 | " <td>0.2731</td>\n", |
||
| 2230 | " <td>0.6346</td>\n", |
||
| 2231 | " <td>-1.1963</td>\n", |
||
| 2232 | " <td>0.1686</td>\n", |
||
| 2233 | " </tr>\n", |
||
| 2234 | " <tr>\n", |
||
| 2235 | " <th>515</th>\n", |
||
| 2236 | " <td>-0.6140</td>\n", |
||
| 2237 | " <td>-0.1379</td>\n", |
||
| 2238 | " <td>0.0493</td>\n", |
||
| 2239 | " <td>0.3641</td>\n", |
||
| 2240 | " <td>-0.3317</td>\n", |
||
| 2241 | " <td>-0.0722</td>\n", |
||
| 2242 | " <td>-0.1809</td>\n", |
||
| 2243 | " <td>-0.1263</td>\n", |
||
| 2244 | " <td>-0.2508</td>\n", |
||
| 2245 | " <td>0.1358</td>\n", |
||
| 2246 | " <td>...</td>\n", |
||
| 2247 | " <td>-1.0456</td>\n", |
||
| 2248 | " <td>0.5245</td>\n", |
||
| 2249 | " <td>-0.1738</td>\n", |
||
| 2250 | " <td>2.4043</td>\n", |
||
| 2251 | " <td>-0.7251</td>\n", |
||
| 2252 | " <td>-1.0053</td>\n", |
||
| 2253 | " <td>0.7014</td>\n", |
||
| 2254 | " <td>0.7755</td>\n", |
||
| 2255 | " <td>-1.0308</td>\n", |
||
| 2256 | " <td>0.6609</td>\n", |
||
| 2257 | " </tr>\n", |
||
| 2258 | " </tbody>\n", |
||
| 2259 | "</table>\n", |
||
| 2260 | "<p>516 rows × 18345 columns</p>\n", |
||
| 2261 | "</div>" |
||
| 2262 | ], |
||
| 2263 | "text/plain": [ |
||
| 2264 | " UBE2Q2P2_rnaseq SSX9_rnaseq CXORF67_rnaseq EFCAB8_rnaseq \\\n", |
||
| 2265 | "0 -0.3291 -0.1379 -0.1805 -0.0869 \n", |
||
| 2266 | "1 -0.8531 -0.1379 -0.1805 -0.2629 \n", |
||
| 2267 | "2 -0.7262 0.3883 0.4908 -0.0666 \n", |
||
| 2268 | "3 -1.0590 -0.1379 -0.1805 -0.0959 \n", |
||
| 2269 | "4 -0.7257 -0.1379 -0.1805 -0.1756 \n", |
||
| 2270 | ".. ... ... ... ... \n", |
||
| 2271 | "511 0.5308 -0.1379 -0.1805 -0.2629 \n", |
||
| 2272 | "512 -0.5021 -0.1379 -0.0120 1.7408 \n", |
||
| 2273 | "513 5.2714 -0.1379 -0.1805 0.1753 \n", |
||
| 2274 | "514 0.6290 -0.1379 0.1131 -0.0667 \n", |
||
| 2275 | "515 -0.6140 -0.1379 0.0493 0.3641 \n", |
||
| 2276 | "\n", |
||
| 2277 | " SDR16C6P_rnaseq EFCAB12_rnaseq A1BG_rnaseq A1CF_rnaseq RBFOX1_rnaseq \\\n", |
||
| 2278 | "0 -0.3317 -0.1661 -0.1483 -0.1371 -0.2260 \n", |
||
| 2279 | "1 -0.3317 -0.2317 -0.5528 -0.1476 -0.2508 \n", |
||
| 2280 | "2 -0.3317 -0.3948 0.0021 -0.1476 -0.2508 \n", |
||
| 2281 | "3 -0.3317 -0.3372 -0.1061 -0.1476 -0.2508 \n", |
||
| 2282 | "4 -0.3317 -0.3778 0.1119 -0.1476 1.2515 \n", |
||
| 2283 | ".. ... ... ... ... ... \n", |
||
| 2284 | "511 -0.3317 -0.2827 -0.6045 -0.1476 -0.2508 \n", |
||
| 2285 | "512 -0.3317 -0.2152 0.7495 1.8708 -0.1178 \n", |
||
| 2286 | "513 -0.3317 -0.2325 0.5863 -0.1476 -0.0185 \n", |
||
| 2287 | "514 1.5316 -0.3634 0.3730 -0.1476 -0.2361 \n", |
||
| 2288 | "515 -0.3317 -0.0722 -0.1809 -0.1263 -0.2508 \n", |
||
| 2289 | "\n", |
||
| 2290 | " GGACT_rnaseq ... ZWINT_rnaseq ZXDA_rnaseq ZXDB_rnaseq ZXDC_rnaseq \\\n", |
||
| 2291 | "0 -0.5346 ... -0.7082 0.6149 0.5725 0.2889 \n", |
||
| 2292 | "1 0.6921 ... 0.9291 0.5118 -0.1673 -0.8006 \n", |
||
| 2293 | "2 -0.0800 ... 0.2957 -0.4399 -0.2751 -0.4668 \n", |
||
| 2294 | "3 -0.5641 ... -0.9962 1.4844 0.9748 0.7481 \n", |
||
| 2295 | "4 -1.0113 ... 1.7870 -0.0462 1.8418 -0.9922 \n", |
||
| 2296 | ".. ... ... ... ... ... ... \n", |
||
| 2297 | "511 -0.2014 ... -0.5331 -0.4205 -0.3773 0.0551 \n", |
||
| 2298 | "512 -1.3502 ... -0.3624 0.0588 -0.1157 1.2831 \n", |
||
| 2299 | "513 -0.3172 ... -0.9598 -1.5823 -1.3015 0.4371 \n", |
||
| 2300 | "514 -1.7106 ... -0.5337 4.2234 0.9716 0.6699 \n", |
||
| 2301 | "515 0.1358 ... -1.0456 0.5245 -0.1738 2.4043 \n", |
||
| 2302 | "\n", |
||
| 2303 | " ZYG11A_rnaseq ZYG11B_rnaseq ZYX_rnaseq ZZEF1_rnaseq ZZZ3_rnaseq \\\n", |
||
| 2304 | "0 -0.5255 -0.2205 -0.7847 -0.2296 -0.0897 \n", |
||
| 2305 | "1 -0.4348 -1.7113 0.7466 -0.1563 -0.9102 \n", |
||
| 2306 | "2 0.1222 0.3555 1.4078 -0.1592 -0.2276 \n", |
||
| 2307 | "3 -0.7049 -0.2617 -0.2934 1.1243 0.0823 \n", |
||
| 2308 | "4 -0.7090 -1.0285 0.6567 -1.0377 -1.1277 \n", |
||
| 2309 | ".. ... ... ... ... ... \n", |
||
| 2310 | "511 -0.5660 -0.5123 0.1254 0.2124 -0.6375 \n", |
||
| 2311 | "512 -0.0555 -0.3620 -0.4242 1.6937 -0.4990 \n", |
||
| 2312 | "513 -0.6739 -1.4417 -0.9613 0.4167 -1.4631 \n", |
||
| 2313 | "514 -0.8134 -0.2453 0.2731 0.6346 -1.1963 \n", |
||
| 2314 | "515 -0.7251 -1.0053 0.7014 0.7755 -1.0308 \n", |
||
| 2315 | "\n", |
||
| 2316 | " TPTEP1_rnaseq \n", |
||
| 2317 | "0 0.1457 \n", |
||
| 2318 | "1 -0.5005 \n", |
||
| 2319 | "2 -0.3931 \n", |
||
| 2320 | "3 0.8831 \n", |
||
| 2321 | "4 -0.5026 \n", |
||
| 2322 | ".. ... \n", |
||
| 2323 | "511 1.4712 \n", |
||
| 2324 | "512 2.2944 \n", |
||
| 2325 | "513 -0.5035 \n", |
||
| 2326 | "514 0.1686 \n", |
||
| 2327 | "515 0.6609 \n", |
||
| 2328 | "\n", |
||
| 2329 | "[516 rows x 18345 columns]" |
||
| 2330 | ] |
||
| 2331 | }, |
||
| 2332 | "execution_count": 25, |
||
| 2333 | "metadata": {}, |
||
| 2334 | "output_type": "execute_result" |
||
| 2335 | } |
||
| 2336 | ], |
||
| 2337 | "source": [ |
||
| 2338 | "slide_df" |
||
| 2339 | ] |
||
| 2340 | }, |
||
| 2341 | { |
||
| 2342 | "cell_type": "code", |
||
| 2343 | "execution_count": 18, |
||
| 2344 | "metadata": {}, |
||
| 2345 | "outputs": [ |
||
| 2346 | { |
||
| 2347 | "name": "stderr", |
||
| 2348 | "output_type": "stream", |
||
| 2349 | "text": [ |
||
| 2350 | "<ipython-input-18-1ae2fdcf544f>:14: DeprecationWarning: The default dtype for empty Series will be 'object' instead of 'float64' in a future version. Specify a dtype explicitly to silence this warning.\n", |
||
| 2351 | " return pd.Series(list(set(s1) & set(s2)))\n", |
||
| 2352 | "/home/mahmoodlab/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3071: DtypeWarning: Columns (2) have mixed types.Specify dtype option on import or set low_memory=False.\n", |
||
| 2353 | " has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n", |
||
| 2354 | "/home/mahmoodlab/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3071: DtypeWarning: Columns (4) have mixed types.Specify dtype option on import or set low_memory=False.\n", |
||
| 2355 | " has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n" |
||
| 2356 | ] |
||
| 2357 | } |
||
| 2358 | ], |
||
| 2359 | "source": [ |
||
| 2360 | "### Snippet for creating genomic signatures\n", |
||
| 2361 | "for fname in os.listdir('./'):\n", |
||
| 2362 | " if fname.endswith('.csv.zip'):\n", |
||
| 2363 | " slide_df = pd.read_csv(fname)\n", |
||
| 2364 | " \n", |
||
| 2365 | " signatures = pd.read_csv('./signatures.csv')\n", |
||
| 2366 | " omic_from_signatures = []\n", |
||
| 2367 | " for col in signatures.columns:\n", |
||
| 2368 | " omic = signatures[col].dropna().unique()\n", |
||
| 2369 | " omic_from_signatures.append(omic)\n", |
||
| 2370 | "\n", |
||
| 2371 | " omic_from_signatures = np.concatenate(omic_from_signatures)\n", |
||
| 2372 | "\n", |
||
| 2373 | " def series_intersection(s1, s2):\n", |
||
| 2374 | " return pd.Series(list(set(s1) & set(s2)))\n", |
||
| 2375 | "\n", |
||
| 2376 | " rnaseq_overlap = np.concatenate([omic_from_signatures+mode for mode in ['_rnaseq']])\n", |
||
| 2377 | " rnaseq_overlap = sorted(series_intersection(rnaseq_overlap, slide_df.columns))\n", |
||
| 2378 | " genomics_mut_cnv = list(slide_df.columns[slide_df.columns.str.contains('_mut|_cnv')])\n", |
||
| 2379 | " \n", |
||
| 2380 | " slide_df[list(slide_df.columns[:9]) + rnaseq_overlap + genomics_mut_cnv].to_csv('../dataset_csv_mutsigdb/%s' % fname)" |
||
| 2381 | ] |
||
| 2382 | }, |
||
| 2383 | { |
||
| 2384 | "cell_type": "code", |
||
| 2385 | "execution_count": null, |
||
| 2386 | "metadata": {}, |
||
| 2387 | "outputs": [], |
||
| 2388 | "source": [] |
||
| 2389 | }, |
||
| 2390 | { |
||
| 2391 | "cell_type": "code", |
||
| 2392 | "execution_count": null, |
||
| 2393 | "metadata": {}, |
||
| 2394 | "outputs": [], |
||
| 2395 | "source": [] |
||
| 2396 | }, |
||
| 2397 | { |
||
| 2398 | "cell_type": "code", |
||
| 2399 | "execution_count": null, |
||
| 2400 | "metadata": {}, |
||
| 2401 | "outputs": [], |
||
| 2402 | "source": [] |
||
| 2403 | }, |
||
| 2404 | { |
||
| 2405 | "cell_type": "code", |
||
| 2406 | "execution_count": null, |
||
| 2407 | "metadata": {}, |
||
| 2408 | "outputs": [], |
||
| 2409 | "source": [] |
||
| 2410 | }, |
||
| 2411 | { |
||
| 2412 | "cell_type": "code", |
||
| 2413 | "execution_count": null, |
||
| 2414 | "metadata": {}, |
||
| 2415 | "outputs": [], |
||
| 2416 | "source": [] |
||
| 2417 | }, |
||
| 2418 | { |
||
| 2419 | "cell_type": "code", |
||
| 2420 | "execution_count": 36, |
||
| 2421 | "metadata": {}, |
||
| 2422 | "outputs": [], |
||
| 2423 | "source": [ |
||
| 2424 | "omic_from_signatures = []\n", |
||
| 2425 | "for col in signatures.columns:\n", |
||
| 2426 | " omic = signatures[col].dropna().unique()\n", |
||
| 2427 | " omic_from_signatures.append(omic)\n", |
||
| 2428 | "\n", |
||
| 2429 | "omic_from_signatures = np.concatenate(omic_from_signatures)\n" |
||
| 2430 | ] |
||
| 2431 | }, |
||
| 2432 | { |
||
| 2433 | "cell_type": "code", |
||
| 2434 | "execution_count": 7, |
||
| 2435 | "metadata": {}, |
||
| 2436 | "outputs": [ |
||
| 2437 | { |
||
| 2438 | "name": "stdout", |
||
| 2439 | "output_type": "stream", |
||
| 2440 | "text": [ |
||
| 2441 | "Tumor Suppressor Genes Embedding Size: 84\n", |
||
| 2442 | "Oncogenes Embedding Size: 314\n", |
||
| 2443 | "Protein Kinases Embedding Size: 498\n", |
||
| 2444 | "Cell Differentiation Markers Embedding Size: 415\n", |
||
| 2445 | "Transcription Factors Embedding Size: 1396\n", |
||
| 2446 | "Cytokines and Growth Factors Embedding Size: 428\n" |
||
| 2447 | ] |
||
| 2448 | } |
||
| 2449 | ], |
||
| 2450 | "source": [ |
||
| 2451 | "\n", |
||
| 2452 | "def series_intersection(s1, s2):\n", |
||
| 2453 | " return pd.Series(list(set(s1) & set(s2)))\n", |
||
| 2454 | "\n", |
||
| 2455 | "sig_names = []\n", |
||
| 2456 | "for col in signatures.columns:\n", |
||
| 2457 | " sig = signatures[col].dropna().unique()\n", |
||
| 2458 | " sig = np.concatenate([sig+mode for mode in ['_mut', '_cnv', '_rnaseq']])\n", |
||
| 2459 | " sig = sorted(series_intersection(sig, genomic_features.columns))\n", |
||
| 2460 | " sig_names.append(sig)\n", |
||
| 2461 | " print('%s Embedding Size: %d' % (col, len(sig)))\n", |
||
| 2462 | "sig_sizes = [len(sig) for sig in sig_names]" |
||
| 2463 | ] |
||
| 2464 | }, |
||
| 2465 | { |
||
| 2466 | "cell_type": "code", |
||
| 2467 | "execution_count": 21, |
||
| 2468 | "metadata": {}, |
||
| 2469 | "outputs": [ |
||
| 2470 | { |
||
| 2471 | "data": { |
||
| 2472 | "text/plain": [ |
||
| 2473 | "['IFNA10_cnv',\n", |
||
| 2474 | " 'IFNA13_cnv',\n", |
||
| 2475 | " 'IFNA14_cnv',\n", |
||
| 2476 | " 'IFNA16_cnv',\n", |
||
| 2477 | " 'IFNA17_cnv',\n", |
||
| 2478 | " 'IFNA1_cnv',\n", |
||
| 2479 | " 'IFNA21_cnv',\n", |
||
| 2480 | " 'IFNA2_cnv',\n", |
||
| 2481 | " 'IFNA4_cnv',\n", |
||
| 2482 | " 'IFNA5_cnv',\n", |
||
| 2483 | " 'IFNA6_cnv',\n", |
||
| 2484 | " 'IFNA7_cnv',\n", |
||
| 2485 | " 'IFNA8_cnv',\n", |
||
| 2486 | " 'IFNB1_cnv',\n", |
||
| 2487 | " 'IFNE_cnv',\n", |
||
| 2488 | " 'IFNW1_cnv',\n", |
||
| 2489 | " 'PDGFRA_cnv']" |
||
| 2490 | ] |
||
| 2491 | }, |
||
| 2492 | "execution_count": 21, |
||
| 2493 | "metadata": {}, |
||
| 2494 | "output_type": "execute_result" |
||
| 2495 | } |
||
| 2496 | ], |
||
| 2497 | "source": [ |
||
| 2498 | "sig" |
||
| 2499 | ] |
||
| 2500 | }, |
||
| 2501 | { |
||
| 2502 | "cell_type": "code", |
||
| 2503 | "execution_count": 434, |
||
| 2504 | "metadata": {}, |
||
| 2505 | "outputs": [ |
||
| 2506 | { |
||
| 2507 | "data": { |
||
| 2508 | "text/html": [ |
||
| 2509 | "<div>\n", |
||
| 2510 | "<style scoped>\n", |
||
| 2511 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 2512 | " vertical-align: middle;\n", |
||
| 2513 | " }\n", |
||
| 2514 | "\n", |
||
| 2515 | " .dataframe tbody tr th {\n", |
||
| 2516 | " vertical-align: top;\n", |
||
| 2517 | " }\n", |
||
| 2518 | "\n", |
||
| 2519 | " .dataframe thead th {\n", |
||
| 2520 | " text-align: right;\n", |
||
| 2521 | " }\n", |
||
| 2522 | "</style>\n", |
||
| 2523 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 2524 | " <thead>\n", |
||
| 2525 | " <tr style=\"text-align: right;\">\n", |
||
| 2526 | " <th></th>\n", |
||
| 2527 | " <th>NDUFS5_cnv</th>\n", |
||
| 2528 | " <th>MACF1_cnv</th>\n", |
||
| 2529 | " <th>RNA5SP44_cnv</th>\n", |
||
| 2530 | " <th>KIAA0754_cnv</th>\n", |
||
| 2531 | " <th>BMP8A_cnv</th>\n", |
||
| 2532 | " <th>PABPC4_cnv</th>\n", |
||
| 2533 | " <th>SNORA55_cnv</th>\n", |
||
| 2534 | " <th>HEYL_cnv</th>\n", |
||
| 2535 | " <th>HPCAL4_cnv</th>\n", |
||
| 2536 | " <th>NT5C1A_cnv</th>\n", |
||
| 2537 | " <th>...</th>\n", |
||
| 2538 | " <th>ZWINT_rnaseq</th>\n", |
||
| 2539 | " <th>ZXDA_rnaseq</th>\n", |
||
| 2540 | " <th>ZXDB_rnaseq</th>\n", |
||
| 2541 | " <th>ZXDC_rnaseq</th>\n", |
||
| 2542 | " <th>ZYG11A_rnaseq</th>\n", |
||
| 2543 | " <th>ZYG11B_rnaseq</th>\n", |
||
| 2544 | " <th>ZYX_rnaseq</th>\n", |
||
| 2545 | " <th>ZZEF1_rnaseq</th>\n", |
||
| 2546 | " <th>ZZZ3_rnaseq</th>\n", |
||
| 2547 | " <th>TPTEP1_rnaseq</th>\n", |
||
| 2548 | " </tr>\n", |
||
| 2549 | " </thead>\n", |
||
| 2550 | " <tbody>\n", |
||
| 2551 | " <tr>\n", |
||
| 2552 | " <th>0</th>\n", |
||
| 2553 | " <td>-1</td>\n", |
||
| 2554 | " <td>-1</td>\n", |
||
| 2555 | " <td>-1</td>\n", |
||
| 2556 | " <td>-1</td>\n", |
||
| 2557 | " <td>-1</td>\n", |
||
| 2558 | " <td>-1</td>\n", |
||
| 2559 | " <td>-1</td>\n", |
||
| 2560 | " <td>-1</td>\n", |
||
| 2561 | " <td>-1</td>\n", |
||
| 2562 | " <td>-1</td>\n", |
||
| 2563 | " <td>...</td>\n", |
||
| 2564 | " <td>-0.8388</td>\n", |
||
| 2565 | " <td>4.1375</td>\n", |
||
| 2566 | " <td>3.9664</td>\n", |
||
| 2567 | " <td>1.8437</td>\n", |
||
| 2568 | " <td>-0.3959</td>\n", |
||
| 2569 | " <td>-0.2561</td>\n", |
||
| 2570 | " <td>-0.2866</td>\n", |
||
| 2571 | " <td>1.8770</td>\n", |
||
| 2572 | " <td>-0.3179</td>\n", |
||
| 2573 | " <td>-0.3633</td>\n", |
||
| 2574 | " </tr>\n", |
||
| 2575 | " <tr>\n", |
||
| 2576 | " <th>1</th>\n", |
||
| 2577 | " <td>2</td>\n", |
||
| 2578 | " <td>2</td>\n", |
||
| 2579 | " <td>2</td>\n", |
||
| 2580 | " <td>2</td>\n", |
||
| 2581 | " <td>2</td>\n", |
||
| 2582 | " <td>2</td>\n", |
||
| 2583 | " <td>2</td>\n", |
||
| 2584 | " <td>2</td>\n", |
||
| 2585 | " <td>2</td>\n", |
||
| 2586 | " <td>2</td>\n", |
||
| 2587 | " <td>...</td>\n", |
||
| 2588 | " <td>-0.1083</td>\n", |
||
| 2589 | " <td>0.3393</td>\n", |
||
| 2590 | " <td>0.2769</td>\n", |
||
| 2591 | " <td>1.7320</td>\n", |
||
| 2592 | " <td>-0.0975</td>\n", |
||
| 2593 | " <td>2.6955</td>\n", |
||
| 2594 | " <td>-0.6741</td>\n", |
||
| 2595 | " <td>1.0323</td>\n", |
||
| 2596 | " <td>1.2766</td>\n", |
||
| 2597 | " <td>-0.3982</td>\n", |
||
| 2598 | " </tr>\n", |
||
| 2599 | " <tr>\n", |
||
| 2600 | " <th>2</th>\n", |
||
| 2601 | " <td>0</td>\n", |
||
| 2602 | " <td>0</td>\n", |
||
| 2603 | " <td>0</td>\n", |
||
| 2604 | " <td>0</td>\n", |
||
| 2605 | " <td>0</td>\n", |
||
| 2606 | " <td>0</td>\n", |
||
| 2607 | " <td>0</td>\n", |
||
| 2608 | " <td>0</td>\n", |
||
| 2609 | " <td>0</td>\n", |
||
| 2610 | " <td>0</td>\n", |
||
| 2611 | " <td>...</td>\n", |
||
| 2612 | " <td>-0.4155</td>\n", |
||
| 2613 | " <td>1.6846</td>\n", |
||
| 2614 | " <td>0.7711</td>\n", |
||
| 2615 | " <td>-0.3061</td>\n", |
||
| 2616 | " <td>-0.5016</td>\n", |
||
| 2617 | " <td>2.8548</td>\n", |
||
| 2618 | " <td>-0.6171</td>\n", |
||
| 2619 | " <td>-0.8608</td>\n", |
||
| 2620 | " <td>-0.0486</td>\n", |
||
| 2621 | " <td>-0.3962</td>\n", |
||
| 2622 | " </tr>\n", |
||
| 2623 | " <tr>\n", |
||
| 2624 | " <th>3</th>\n", |
||
| 2625 | " <td>0</td>\n", |
||
| 2626 | " <td>0</td>\n", |
||
| 2627 | " <td>0</td>\n", |
||
| 2628 | " <td>0</td>\n", |
||
| 2629 | " <td>0</td>\n", |
||
| 2630 | " <td>0</td>\n", |
||
| 2631 | " <td>0</td>\n", |
||
| 2632 | " <td>0</td>\n", |
||
| 2633 | " <td>0</td>\n", |
||
| 2634 | " <td>0</td>\n", |
||
| 2635 | " <td>...</td>\n", |
||
| 2636 | " <td>-0.8143</td>\n", |
||
| 2637 | " <td>0.8344</td>\n", |
||
| 2638 | " <td>1.5075</td>\n", |
||
| 2639 | " <td>3.6068</td>\n", |
||
| 2640 | " <td>-0.5004</td>\n", |
||
| 2641 | " <td>-0.0747</td>\n", |
||
| 2642 | " <td>-0.2185</td>\n", |
||
| 2643 | " <td>-0.4379</td>\n", |
||
| 2644 | " <td>1.6913</td>\n", |
||
| 2645 | " <td>1.7748</td>\n", |
||
| 2646 | " </tr>\n", |
||
| 2647 | " <tr>\n", |
||
| 2648 | " <th>4</th>\n", |
||
| 2649 | " <td>0</td>\n", |
||
| 2650 | " <td>0</td>\n", |
||
| 2651 | " <td>0</td>\n", |
||
| 2652 | " <td>0</td>\n", |
||
| 2653 | " <td>0</td>\n", |
||
| 2654 | " <td>0</td>\n", |
||
| 2655 | " <td>0</td>\n", |
||
| 2656 | " <td>0</td>\n", |
||
| 2657 | " <td>0</td>\n", |
||
| 2658 | " <td>0</td>\n", |
||
| 2659 | " <td>...</td>\n", |
||
| 2660 | " <td>0.0983</td>\n", |
||
| 2661 | " <td>-0.7908</td>\n", |
||
| 2662 | " <td>-0.0053</td>\n", |
||
| 2663 | " <td>-0.0643</td>\n", |
||
| 2664 | " <td>-0.3706</td>\n", |
||
| 2665 | " <td>0.3870</td>\n", |
||
| 2666 | " <td>-0.5589</td>\n", |
||
| 2667 | " <td>-0.5979</td>\n", |
||
| 2668 | " <td>0.0047</td>\n", |
||
| 2669 | " <td>-0.3548</td>\n", |
||
| 2670 | " </tr>\n", |
||
| 2671 | " <tr>\n", |
||
| 2672 | " <th>...</th>\n", |
||
| 2673 | " <td>...</td>\n", |
||
| 2674 | " <td>...</td>\n", |
||
| 2675 | " <td>...</td>\n", |
||
| 2676 | " <td>...</td>\n", |
||
| 2677 | " <td>...</td>\n", |
||
| 2678 | " <td>...</td>\n", |
||
| 2679 | " <td>...</td>\n", |
||
| 2680 | " <td>...</td>\n", |
||
| 2681 | " <td>...</td>\n", |
||
| 2682 | " <td>...</td>\n", |
||
| 2683 | " <td>...</td>\n", |
||
| 2684 | " <td>...</td>\n", |
||
| 2685 | " <td>...</td>\n", |
||
| 2686 | " <td>...</td>\n", |
||
| 2687 | " <td>...</td>\n", |
||
| 2688 | " <td>...</td>\n", |
||
| 2689 | " <td>...</td>\n", |
||
| 2690 | " <td>...</td>\n", |
||
| 2691 | " <td>...</td>\n", |
||
| 2692 | " <td>...</td>\n", |
||
| 2693 | " <td>...</td>\n", |
||
| 2694 | " </tr>\n", |
||
| 2695 | " <tr>\n", |
||
| 2696 | " <th>368</th>\n", |
||
| 2697 | " <td>2</td>\n", |
||
| 2698 | " <td>2</td>\n", |
||
| 2699 | " <td>2</td>\n", |
||
| 2700 | " <td>2</td>\n", |
||
| 2701 | " <td>2</td>\n", |
||
| 2702 | " <td>2</td>\n", |
||
| 2703 | " <td>2</td>\n", |
||
| 2704 | " <td>2</td>\n", |
||
| 2705 | " <td>2</td>\n", |
||
| 2706 | " <td>2</td>\n", |
||
| 2707 | " <td>...</td>\n", |
||
| 2708 | " <td>-0.0291</td>\n", |
||
| 2709 | " <td>-0.1058</td>\n", |
||
| 2710 | " <td>-0.6721</td>\n", |
||
| 2711 | " <td>0.2802</td>\n", |
||
| 2712 | " <td>1.9504</td>\n", |
||
| 2713 | " <td>-0.8784</td>\n", |
||
| 2714 | " <td>0.9506</td>\n", |
||
| 2715 | " <td>0.0607</td>\n", |
||
| 2716 | " <td>1.1883</td>\n", |
||
| 2717 | " <td>-0.3521</td>\n", |
||
| 2718 | " </tr>\n", |
||
| 2719 | " <tr>\n", |
||
| 2720 | " <th>369</th>\n", |
||
| 2721 | " <td>0</td>\n", |
||
| 2722 | " <td>0</td>\n", |
||
| 2723 | " <td>0</td>\n", |
||
| 2724 | " <td>0</td>\n", |
||
| 2725 | " <td>0</td>\n", |
||
| 2726 | " <td>0</td>\n", |
||
| 2727 | " <td>0</td>\n", |
||
| 2728 | " <td>0</td>\n", |
||
| 2729 | " <td>0</td>\n", |
||
| 2730 | " <td>0</td>\n", |
||
| 2731 | " <td>...</td>\n", |
||
| 2732 | " <td>0.0497</td>\n", |
||
| 2733 | " <td>0.3673</td>\n", |
||
| 2734 | " <td>-0.2208</td>\n", |
||
| 2735 | " <td>0.3034</td>\n", |
||
| 2736 | " <td>3.2580</td>\n", |
||
| 2737 | " <td>-0.2089</td>\n", |
||
| 2738 | " <td>1.6053</td>\n", |
||
| 2739 | " <td>-0.8746</td>\n", |
||
| 2740 | " <td>-0.4491</td>\n", |
||
| 2741 | " <td>-0.3450</td>\n", |
||
| 2742 | " </tr>\n", |
||
| 2743 | " <tr>\n", |
||
| 2744 | " <th>370</th>\n", |
||
| 2745 | " <td>1</td>\n", |
||
| 2746 | " <td>1</td>\n", |
||
| 2747 | " <td>1</td>\n", |
||
| 2748 | " <td>1</td>\n", |
||
| 2749 | " <td>1</td>\n", |
||
| 2750 | " <td>1</td>\n", |
||
| 2751 | " <td>1</td>\n", |
||
| 2752 | " <td>1</td>\n", |
||
| 2753 | " <td>1</td>\n", |
||
| 2754 | " <td>1</td>\n", |
||
| 2755 | " <td>...</td>\n", |
||
| 2756 | " <td>0.3822</td>\n", |
||
| 2757 | " <td>-0.7003</td>\n", |
||
| 2758 | " <td>-0.7661</td>\n", |
||
| 2759 | " <td>-1.7035</td>\n", |
||
| 2760 | " <td>-0.5423</td>\n", |
||
| 2761 | " <td>-0.3488</td>\n", |
||
| 2762 | " <td>1.3713</td>\n", |
||
| 2763 | " <td>-0.4365</td>\n", |
||
| 2764 | " <td>2.3456</td>\n", |
||
| 2765 | " <td>-0.3866</td>\n", |
||
| 2766 | " </tr>\n", |
||
| 2767 | " <tr>\n", |
||
| 2768 | " <th>371</th>\n", |
||
| 2769 | " <td>0</td>\n", |
||
| 2770 | " <td>0</td>\n", |
||
| 2771 | " <td>0</td>\n", |
||
| 2772 | " <td>0</td>\n", |
||
| 2773 | " <td>0</td>\n", |
||
| 2774 | " <td>0</td>\n", |
||
| 2775 | " <td>0</td>\n", |
||
| 2776 | " <td>0</td>\n", |
||
| 2777 | " <td>0</td>\n", |
||
| 2778 | " <td>0</td>\n", |
||
| 2779 | " <td>...</td>\n", |
||
| 2780 | " <td>-0.6853</td>\n", |
||
| 2781 | " <td>-1.0240</td>\n", |
||
| 2782 | " <td>-1.2890</td>\n", |
||
| 2783 | " <td>-1.5666</td>\n", |
||
| 2784 | " <td>-0.1270</td>\n", |
||
| 2785 | " <td>-1.4662</td>\n", |
||
| 2786 | " <td>0.3981</td>\n", |
||
| 2787 | " <td>-0.5976</td>\n", |
||
| 2788 | " <td>-1.3822</td>\n", |
||
| 2789 | " <td>-0.4157</td>\n", |
||
| 2790 | " </tr>\n", |
||
| 2791 | " <tr>\n", |
||
| 2792 | " <th>372</th>\n", |
||
| 2793 | " <td>0</td>\n", |
||
| 2794 | " <td>0</td>\n", |
||
| 2795 | " <td>0</td>\n", |
||
| 2796 | " <td>0</td>\n", |
||
| 2797 | " <td>0</td>\n", |
||
| 2798 | " <td>0</td>\n", |
||
| 2799 | " <td>0</td>\n", |
||
| 2800 | " <td>0</td>\n", |
||
| 2801 | " <td>0</td>\n", |
||
| 2802 | " <td>0</td>\n", |
||
| 2803 | " <td>...</td>\n", |
||
| 2804 | " <td>0.0517</td>\n", |
||
| 2805 | " <td>-0.3570</td>\n", |
||
| 2806 | " <td>-0.4843</td>\n", |
||
| 2807 | " <td>-0.3792</td>\n", |
||
| 2808 | " <td>-0.1964</td>\n", |
||
| 2809 | " <td>0.4200</td>\n", |
||
| 2810 | " <td>3.2547</td>\n", |
||
| 2811 | " <td>-0.1232</td>\n", |
||
| 2812 | " <td>3.4519</td>\n", |
||
| 2813 | " <td>-0.1962</td>\n", |
||
| 2814 | " </tr>\n", |
||
| 2815 | " </tbody>\n", |
||
| 2816 | "</table>\n", |
||
| 2817 | "<p>373 rows × 20395 columns</p>\n", |
||
| 2818 | "</div>" |
||
| 2819 | ], |
||
| 2820 | "text/plain": [ |
||
| 2821 | " NDUFS5_cnv MACF1_cnv RNA5SP44_cnv KIAA0754_cnv BMP8A_cnv PABPC4_cnv \\\n", |
||
| 2822 | "0 -1 -1 -1 -1 -1 -1 \n", |
||
| 2823 | "1 2 2 2 2 2 2 \n", |
||
| 2824 | "2 0 0 0 0 0 0 \n", |
||
| 2825 | "3 0 0 0 0 0 0 \n", |
||
| 2826 | "4 0 0 0 0 0 0 \n", |
||
| 2827 | ".. ... ... ... ... ... ... \n", |
||
| 2828 | "368 2 2 2 2 2 2 \n", |
||
| 2829 | "369 0 0 0 0 0 0 \n", |
||
| 2830 | "370 1 1 1 1 1 1 \n", |
||
| 2831 | "371 0 0 0 0 0 0 \n", |
||
| 2832 | "372 0 0 0 0 0 0 \n", |
||
| 2833 | "\n", |
||
| 2834 | " SNORA55_cnv HEYL_cnv HPCAL4_cnv NT5C1A_cnv ... ZWINT_rnaseq \\\n", |
||
| 2835 | "0 -1 -1 -1 -1 ... -0.8388 \n", |
||
| 2836 | "1 2 2 2 2 ... -0.1083 \n", |
||
| 2837 | "2 0 0 0 0 ... -0.4155 \n", |
||
| 2838 | "3 0 0 0 0 ... -0.8143 \n", |
||
| 2839 | "4 0 0 0 0 ... 0.0983 \n", |
||
| 2840 | ".. ... ... ... ... ... ... \n", |
||
| 2841 | "368 2 2 2 2 ... -0.0291 \n", |
||
| 2842 | "369 0 0 0 0 ... 0.0497 \n", |
||
| 2843 | "370 1 1 1 1 ... 0.3822 \n", |
||
| 2844 | "371 0 0 0 0 ... -0.6853 \n", |
||
| 2845 | "372 0 0 0 0 ... 0.0517 \n", |
||
| 2846 | "\n", |
||
| 2847 | " ZXDA_rnaseq ZXDB_rnaseq ZXDC_rnaseq ZYG11A_rnaseq ZYG11B_rnaseq \\\n", |
||
| 2848 | "0 4.1375 3.9664 1.8437 -0.3959 -0.2561 \n", |
||
| 2849 | "1 0.3393 0.2769 1.7320 -0.0975 2.6955 \n", |
||
| 2850 | "2 1.6846 0.7711 -0.3061 -0.5016 2.8548 \n", |
||
| 2851 | "3 0.8344 1.5075 3.6068 -0.5004 -0.0747 \n", |
||
| 2852 | "4 -0.7908 -0.0053 -0.0643 -0.3706 0.3870 \n", |
||
| 2853 | ".. ... ... ... ... ... \n", |
||
| 2854 | "368 -0.1058 -0.6721 0.2802 1.9504 -0.8784 \n", |
||
| 2855 | "369 0.3673 -0.2208 0.3034 3.2580 -0.2089 \n", |
||
| 2856 | "370 -0.7003 -0.7661 -1.7035 -0.5423 -0.3488 \n", |
||
| 2857 | "371 -1.0240 -1.2890 -1.5666 -0.1270 -1.4662 \n", |
||
| 2858 | "372 -0.3570 -0.4843 -0.3792 -0.1964 0.4200 \n", |
||
| 2859 | "\n", |
||
| 2860 | " ZYX_rnaseq ZZEF1_rnaseq ZZZ3_rnaseq TPTEP1_rnaseq \n", |
||
| 2861 | "0 -0.2866 1.8770 -0.3179 -0.3633 \n", |
||
| 2862 | "1 -0.6741 1.0323 1.2766 -0.3982 \n", |
||
| 2863 | "2 -0.6171 -0.8608 -0.0486 -0.3962 \n", |
||
| 2864 | "3 -0.2185 -0.4379 1.6913 1.7748 \n", |
||
| 2865 | "4 -0.5589 -0.5979 0.0047 -0.3548 \n", |
||
| 2866 | ".. ... ... ... ... \n", |
||
| 2867 | "368 0.9506 0.0607 1.1883 -0.3521 \n", |
||
| 2868 | "369 1.6053 -0.8746 -0.4491 -0.3450 \n", |
||
| 2869 | "370 1.3713 -0.4365 2.3456 -0.3866 \n", |
||
| 2870 | "371 0.3981 -0.5976 -1.3822 -0.4157 \n", |
||
| 2871 | "372 3.2547 -0.1232 3.4519 -0.1962 \n", |
||
| 2872 | "\n", |
||
| 2873 | "[373 rows x 20395 columns]" |
||
| 2874 | ] |
||
| 2875 | }, |
||
| 2876 | "execution_count": 434, |
||
| 2877 | "metadata": {}, |
||
| 2878 | "output_type": "execute_result" |
||
| 2879 | } |
||
| 2880 | ], |
||
| 2881 | "source": [ |
||
| 2882 | "genomic_features" |
||
| 2883 | ] |
||
| 2884 | }, |
||
| 2885 | { |
||
| 2886 | "cell_type": "code", |
||
| 2887 | "execution_count": 2, |
||
| 2888 | "metadata": {}, |
||
| 2889 | "outputs": [], |
||
| 2890 | "source": [ |
||
| 2891 | "import torch\n", |
||
| 2892 | "import torch.nn as nn\n", |
||
| 2893 | "import torch.nn.functional as F\n", |
||
| 2894 | "import pdb\n", |
||
| 2895 | "import numpy as np\n", |
||
| 2896 | "\n", |
||
| 2897 | "class MIL_Sum_FC_surv(nn.Module):\n", |
||
| 2898 | " def __init__(self, size_arg = \"small\", dropout=0.25, n_classes=4):\n", |
||
| 2899 | " super(MIL_Sum_FC_surv, self).__init__()\n", |
||
| 2900 | "\n", |
||
| 2901 | " self.size_dict = {\"small\": [1024, 512, 256], \"big\": [1024, 512, 384]}\n", |
||
| 2902 | " size = self.size_dict[size_arg]\n", |
||
| 2903 | " self.phi = nn.Sequential(*[nn.Linear(size[0], size[1]), nn.ReLU(), nn.Dropout(dropout)])\n", |
||
| 2904 | " self.rho = nn.Sequential(*[nn.Linear(size[1], size[2]), nn.ReLU(), nn.Dropout(dropout)])\n", |
||
| 2905 | " self.classifier = nn.Linear(size[2], n_classes)\n", |
||
| 2906 | "\n", |
||
| 2907 | " def relocate(self):\n", |
||
| 2908 | " device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n", |
||
| 2909 | " if torch.cuda.device_count() >= 1:\n", |
||
| 2910 | " device_ids = list(range(torch.cuda.device_count()))\n", |
||
| 2911 | " self.phi = nn.DataParallel(self.phi, device_ids=device_ids).to('cuda:0')\n", |
||
| 2912 | "\n", |
||
| 2913 | " self.rho = self.rho.to(device)\n", |
||
| 2914 | " self.classifier = self.classifier.to(device)\n", |
||
| 2915 | "\n", |
||
| 2916 | " def forward(self, **kwargs):\n", |
||
| 2917 | " h = kwargs['x_path']\n", |
||
| 2918 | "\n", |
||
| 2919 | " h = self.phi(h).sum(axis=0)\n", |
||
| 2920 | " h = self.rho(h)\n", |
||
| 2921 | " logits = self.classifier(h).unsqueeze(0)\n", |
||
| 2922 | " Y_hat = torch.topk(logits, 1, dim = 1)[1]\n", |
||
| 2923 | " hazards = torch.sigmoid(logits)\n", |
||
| 2924 | " S = torch.cumprod(1 - hazards, dim=1)\n", |
||
| 2925 | " \n", |
||
| 2926 | " return hazards, S, Y_hat, None, None\n", |
||
| 2927 | "\n", |
||
| 2928 | "from os.path import join\n", |
||
| 2929 | "from collections import OrderedDict\n", |
||
| 2930 | "\n", |
||
| 2931 | "import torch\n", |
||
| 2932 | "import torch.nn as nn\n", |
||
| 2933 | "import torch.nn.functional as F\n", |
||
| 2934 | "import pdb\n", |
||
| 2935 | "import numpy as np\n", |
||
| 2936 | "\n", |
||
| 2937 | "\"\"\"\n", |
||
| 2938 | "A Modified Implementation of Deep Attention MIL\n", |
||
| 2939 | "\"\"\"\n", |
||
| 2940 | "\n", |
||
| 2941 | "\n", |
||
| 2942 | "\"\"\"\n", |
||
| 2943 | "Attention Network without Gating (2 fc layers)\n", |
||
| 2944 | "args:\n", |
||
| 2945 | " L: input feature dimension\n", |
||
| 2946 | " D: hidden layer dimension\n", |
||
| 2947 | " dropout: whether to use dropout (p = 0.25)\n", |
||
| 2948 | " n_classes: number of classes (experimental usage for multiclass MIL)\n", |
||
| 2949 | "\"\"\"\n", |
||
| 2950 | "class Attn_Net(nn.Module):\n", |
||
| 2951 | "\n", |
||
| 2952 | " def __init__(self, L = 1024, D = 256, dropout = False, n_classes = 1):\n", |
||
| 2953 | " super(Attn_Net, self).__init__()\n", |
||
| 2954 | " self.module = [\n", |
||
| 2955 | " nn.Linear(L, D),\n", |
||
| 2956 | " nn.Tanh()]\n", |
||
| 2957 | "\n", |
||
| 2958 | " if dropout:\n", |
||
| 2959 | " self.module.append(nn.Dropout(0.25))\n", |
||
| 2960 | "\n", |
||
| 2961 | " self.module.append(nn.Linear(D, n_classes))\n", |
||
| 2962 | " \n", |
||
| 2963 | " self.module = nn.Sequential(*self.module)\n", |
||
| 2964 | " \n", |
||
| 2965 | " def forward(self, x):\n", |
||
| 2966 | " return self.module(x), x # N x n_classes\n", |
||
| 2967 | "\n", |
||
| 2968 | "\"\"\"\n", |
||
| 2969 | "Attention Network with Sigmoid Gating (3 fc layers)\n", |
||
| 2970 | "args:\n", |
||
| 2971 | " L: input feature dimension\n", |
||
| 2972 | " D: hidden layer dimension\n", |
||
| 2973 | " dropout: whether to use dropout (p = 0.25)\n", |
||
| 2974 | " n_classes: number of classes (experimental usage for multiclass MIL)\n", |
||
| 2975 | "\"\"\"\n", |
||
| 2976 | "class Attn_Net_Gated(nn.Module):\n", |
||
| 2977 | "\n", |
||
| 2978 | " def __init__(self, L = 1024, D = 256, dropout = False, n_classes = 1):\n", |
||
| 2979 | " super(Attn_Net_Gated, self).__init__()\n", |
||
| 2980 | " self.attention_a = [\n", |
||
| 2981 | " nn.Linear(L, D),\n", |
||
| 2982 | " nn.Tanh()]\n", |
||
| 2983 | " \n", |
||
| 2984 | " self.attention_b = [nn.Linear(L, D),\n", |
||
| 2985 | " nn.Sigmoid()]\n", |
||
| 2986 | " if dropout:\n", |
||
| 2987 | " self.attention_a.append(nn.Dropout(0.25))\n", |
||
| 2988 | " self.attention_b.append(nn.Dropout(0.25))\n", |
||
| 2989 | "\n", |
||
| 2990 | " self.attention_a = nn.Sequential(*self.attention_a)\n", |
||
| 2991 | " self.attention_b = nn.Sequential(*self.attention_b)\n", |
||
| 2992 | " \n", |
||
| 2993 | " self.attention_c = nn.Linear(D, n_classes)\n", |
||
| 2994 | "\n", |
||
| 2995 | " def forward(self, x):\n", |
||
| 2996 | " a = self.attention_a(x)\n", |
||
| 2997 | " b = self.attention_b(x)\n", |
||
| 2998 | " A = a.mul(b)\n", |
||
| 2999 | " A = self.attention_c(A) # N x n_classes\n", |
||
| 3000 | " return A, x\n", |
||
| 3001 | " \n", |
||
| 3002 | "class MIL_Cluster_FC_surv(nn.Module):\n", |
||
| 3003 | " def __init__(self, num_clusters=10, size_arg = \"small\", dropout=0.25, n_classes=4):\n", |
||
| 3004 | " super(MIL_Cluster_FC_surv, self).__init__()\n", |
||
| 3005 | " self.size_dict = {\"small\": [1024, 512, 256], \"big\": [1024, 512, 384]}\n", |
||
| 3006 | " self.num_clusters = num_clusters\n", |
||
| 3007 | " \n", |
||
| 3008 | " ### Phenotype Learning\n", |
||
| 3009 | " size = self.size_dict[size_arg]\n", |
||
| 3010 | " phis = []\n", |
||
| 3011 | " for phenotype_i in range(num_clusters):\n", |
||
| 3012 | " phi = [nn.Linear(size[0], size[1]), nn.ReLU(), nn.Dropout(dropout),\n", |
||
| 3013 | " nn.Linear(size[1], size[1]), nn.ReLU(), nn.Dropout(dropout)]\n", |
||
| 3014 | " phis.append(nn.Sequential(*phi))\n", |
||
| 3015 | " self.phis = nn.ModuleList(phis)\n", |
||
| 3016 | " self.pool1d = nn.AdaptiveAvgPool1d(1)\n", |
||
| 3017 | " \n", |
||
| 3018 | " \n", |
||
| 3019 | " ### WSI Attention MIL Construction\n", |
||
| 3020 | " fc = [nn.Linear(size[1], size[1]), nn.ReLU()]\n", |
||
| 3021 | " fc.append(nn.Dropout(0.25))\n", |
||
| 3022 | " attention_net = Attn_Net_Gated(L=size[1], D=size[2], dropout=dropout, n_classes=1)\n", |
||
| 3023 | " fc.append(attention_net)\n", |
||
| 3024 | " self.attention_net = nn.Sequential(*fc)\n", |
||
| 3025 | "\n", |
||
| 3026 | " \n", |
||
| 3027 | " self.rho = nn.Sequential(*[nn.Linear(size[1], size[2]), nn.ReLU(), nn.Dropout(dropout)])\n", |
||
| 3028 | " self.classifier = nn.Linear(size[2], n_classes)\n", |
||
| 3029 | "\n", |
||
| 3030 | " def relocate(self):\n", |
||
| 3031 | " device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n", |
||
| 3032 | " if torch.cuda.device_count() >= 1:\n", |
||
| 3033 | " device_ids = list(range(torch.cuda.device_count()))\n", |
||
| 3034 | " self.phis = nn.DataParallel(self.phi, device_ids=device_ids).to('cuda:0')\n", |
||
| 3035 | "\n", |
||
| 3036 | " self.rho = self.rho.to(device)\n", |
||
| 3037 | " self.classifier = self.classifier.to(device)\n", |
||
| 3038 | "\n", |
||
| 3039 | " def forward(self, **kwargs):\n", |
||
| 3040 | " x_path = kwargs['x_path']\n", |
||
| 3041 | " ### Phenotyping\n", |
||
| 3042 | " h_phenotypes = []\n", |
||
| 3043 | " from sklearn.cluster import KMeans\n", |
||
| 3044 | " kmeans = KMeans(n_clusters=self.num_clusters, random_state=2021).fit(X)\n", |
||
| 3045 | " #cluster_ids_x, cluster_centers = kmeans(X=x_path, num_clusters=self.num_clusters, distance='euclidean', device=torch.device('cpu'))\n", |
||
| 3046 | " cluster_ids_x = KMeans(n_clusters=10, random_state=2021, max_iter=20).fit_predict(x_path)\n", |
||
| 3047 | " for i in range(self.num_clusters):\n", |
||
| 3048 | " h_phenotypes_i = self.phis[i](x_path[cluster_ids_x==i])\n", |
||
| 3049 | " h_phenotypes.append(self.pool1d(h_phenotypes_i.T.unsqueeze(0)).squeeze(2))\n", |
||
| 3050 | " h_phenotypes = torch.stack(h_phenotypes, dim=1).squeeze(0)\n", |
||
| 3051 | "\n", |
||
| 3052 | "\n", |
||
| 3053 | " ### Attention MIL\n", |
||
| 3054 | " A, h = self.attention_net(h_phenotypes) \n", |
||
| 3055 | " A = torch.transpose(A, 1, 0)\n", |
||
| 3056 | " if 'attention_only' in kwargs.keys():\n", |
||
| 3057 | " if kwargs['attention_only']:\n", |
||
| 3058 | " return A\n", |
||
| 3059 | " A_raw = A \n", |
||
| 3060 | " A = F.softmax(A, dim=1) \n", |
||
| 3061 | " h = torch.mm(A, h_phenotypes)\n", |
||
| 3062 | "\n", |
||
| 3063 | " \n", |
||
| 3064 | " h = self.rho(h)\n", |
||
| 3065 | " logits = self.classifier(h).unsqueeze(0)\n", |
||
| 3066 | " Y_hat = torch.topk(logits, 1, dim = 1)[1]\n", |
||
| 3067 | " hazards = torch.sigmoid(logits)\n", |
||
| 3068 | " S = torch.cumprod(1 - hazards, dim=1)\n", |
||
| 3069 | " \n", |
||
| 3070 | " return hazards, S, Y_hat, None, None" |
||
| 3071 | ] |
||
| 3072 | }, |
||
| 3073 | { |
||
| 3074 | "cell_type": "code", |
||
| 3075 | "execution_count": 15, |
||
| 3076 | "metadata": {}, |
||
| 3077 | "outputs": [], |
||
| 3078 | "source": [ |
||
| 3079 | "x_path = torch.randint(10, size=(500, 1024)).type(torch.cuda.FloatTensor)\n" |
||
| 3080 | ] |
||
| 3081 | }, |
||
| 3082 | { |
||
| 3083 | "cell_type": "code", |
||
| 3084 | "execution_count": 17, |
||
| 3085 | "metadata": {}, |
||
| 3086 | "outputs": [], |
||
| 3087 | "source": [ |
||
| 3088 | "from sklearn.cluster import KMeans\n", |
||
| 3089 | "kmeans = KMeans(n_clusters=10, random_state=2021, max_iter=20).fit_predict(x_path.cpu())" |
||
| 3090 | ] |
||
| 3091 | }, |
||
| 3092 | { |
||
| 3093 | "cell_type": "code", |
||
| 3094 | "execution_count": 18, |
||
| 3095 | "metadata": {}, |
||
| 3096 | "outputs": [ |
||
| 3097 | { |
||
| 3098 | "data": { |
||
| 3099 | "text/plain": [ |
||
| 3100 | "array([5, 5, 3, 5, 8, 4, 8, 7, 5, 4, 9, 1, 9, 1, 6, 1, 1, 0, 5, 0, 4, 3,\n", |
||
| 3101 | " 0, 6, 3, 1, 0, 7, 9, 8, 0, 5, 5, 3, 0, 1, 5, 1, 0, 6, 6, 4, 1, 5,\n", |
||
| 3102 | " 3, 0, 1, 0, 8, 5, 1, 8, 1, 0, 5, 0, 2, 5, 6, 5, 0, 0, 5, 1, 2, 7,\n", |
||
| 3103 | " 4, 6, 5, 3, 0, 7, 9, 1, 3, 4, 4, 5, 7, 9, 9, 5, 0, 1, 9, 1, 2, 0,\n", |
||
| 3104 | " 6, 3, 1, 1, 2, 4, 0, 5, 1, 1, 1, 0, 0, 9, 8, 1, 5, 5, 0, 9, 2, 3,\n", |
||
| 3105 | " 7, 0, 1, 6, 7, 5, 3, 5, 0, 1, 6, 1, 6, 2, 8, 7, 6, 1, 6, 2, 5, 0,\n", |
||
| 3106 | " 1, 6, 0, 9, 2, 1, 0, 1, 7, 7, 6, 1, 6, 0, 3, 4, 1, 3, 2, 4, 4, 5,\n", |
||
| 3107 | " 4, 1, 1, 9, 6, 0, 3, 6, 4, 8, 7, 9, 6, 5, 5, 9, 0, 6, 0, 1, 9, 2,\n", |
||
| 3108 | " 3, 5, 1, 9, 6, 1, 0, 6, 6, 0, 0, 6, 7, 1, 6, 1, 1, 1, 4, 0, 2, 1,\n", |
||
| 3109 | " 9, 5, 7, 5, 9, 0, 1, 0, 6, 2, 2, 1, 1, 5, 3, 5, 3, 6, 5, 6, 9, 5,\n", |
||
| 3110 | " 2, 2, 2, 6, 0, 0, 0, 5, 2, 6, 6, 0, 2, 5, 1, 9, 2, 4, 4, 0, 4, 7,\n", |
||
| 3111 | " 4, 1, 1, 3, 6, 0, 1, 2, 4, 0, 8, 1, 8, 5, 5, 7, 4, 1, 6, 1, 0, 8,\n", |
||
| 3112 | " 6, 1, 1, 4, 8, 7, 5, 2, 3, 0, 2, 9, 5, 6, 4, 3, 6, 5, 5, 4, 6, 6,\n", |
||
| 3113 | " 0, 1, 5, 1, 1, 1, 1, 9, 5, 7, 3, 0, 2, 4, 0, 5, 4, 0, 5, 0, 6, 0,\n", |
||
| 3114 | " 3, 1, 4, 6, 3, 7, 1, 6, 7, 0, 1, 4, 6, 1, 6, 0, 6, 0, 5, 9, 1, 1,\n", |
||
| 3115 | " 3, 1, 5, 6, 1, 6, 6, 8, 2, 0, 7, 9, 9, 6, 0, 6, 2, 6, 8, 0, 8, 5,\n", |
||
| 3116 | " 1, 3, 1, 9, 2, 3, 5, 8, 2, 5, 6, 6, 5, 2, 9, 0, 1, 8, 5, 9, 5, 1,\n", |
||
| 3117 | " 0, 1, 0, 8, 6, 1, 7, 2, 8, 3, 1, 6, 2, 2, 1, 6, 0, 2, 6, 1, 1, 4,\n", |
||
| 3118 | " 5, 6, 4, 0, 5, 0, 9, 0, 4, 8, 0, 7, 6, 5, 5, 0, 4, 1, 1, 2, 2, 0,\n", |
||
| 3119 | " 0, 6, 4, 0, 7, 7, 2, 3, 1, 4, 7, 9, 4, 7, 2, 4, 5, 6, 4, 5, 7, 9,\n", |
||
| 3120 | " 8, 0, 6, 2, 0, 6, 6, 3, 5, 4, 4, 0, 1, 0, 5, 3, 1, 6, 0, 7, 4, 1,\n", |
||
| 3121 | " 6, 3, 6, 0, 4, 1, 5, 7, 3, 1, 4, 8, 0, 7, 0, 6, 1, 1, 0, 1, 5, 1,\n", |
||
| 3122 | " 2, 3, 2, 3, 8, 8, 4, 6, 5, 6, 1, 0, 7, 6, 4, 4], dtype=int32)" |
||
| 3123 | ] |
||
| 3124 | }, |
||
| 3125 | "execution_count": 18, |
||
| 3126 | "metadata": {}, |
||
| 3127 | "output_type": "execute_result" |
||
| 3128 | } |
||
| 3129 | ], |
||
| 3130 | "source": [ |
||
| 3131 | "kmeans" |
||
| 3132 | ] |
||
| 3133 | }, |
||
| 3134 | { |
||
| 3135 | "cell_type": "code", |
||
| 3136 | "execution_count": 2, |
||
| 3137 | "metadata": {}, |
||
| 3138 | "outputs": [ |
||
| 3139 | { |
||
| 3140 | "data": { |
||
| 3141 | "text/plain": [ |
||
| 3142 | "(tensor([[0.9992, 0.0000, 0.0000, 1.0000]], grad_fn=<SigmoidBackward>),\n", |
||
| 3143 | " tensor([[0.0008, 0.0008, 0.0008, 0.0000]], grad_fn=<CumprodBackward>),\n", |
||
| 3144 | " tensor([[3]]),\n", |
||
| 3145 | " None,\n", |
||
| 3146 | " None)" |
||
| 3147 | ] |
||
| 3148 | }, |
||
| 3149 | "execution_count": 2, |
||
| 3150 | "metadata": {}, |
||
| 3151 | "output_type": "execute_result" |
||
| 3152 | } |
||
| 3153 | ], |
||
| 3154 | "source": [ |
||
| 3155 | "x_path = torch.randint(10, size=(500, 1024)).type(torch.FloatTensor)\n", |
||
| 3156 | "model = MIL_Sum_FC_surv()\n", |
||
| 3157 | "model.forward(x_path=x_path)" |
||
| 3158 | ] |
||
| 3159 | }, |
||
| 3160 | { |
||
| 3161 | "cell_type": "code", |
||
| 3162 | "execution_count": 3, |
||
| 3163 | "metadata": {}, |
||
| 3164 | "outputs": [ |
||
| 3165 | { |
||
| 3166 | "data": { |
||
| 3167 | "text/plain": [ |
||
| 3168 | "(tensor([[4.2595e-07, 1.0000e+00, 0.0000e+00, 7.2488e-12]],\n", |
||
| 3169 | " grad_fn=<SigmoidBackward>),\n", |
||
| 3170 | " tensor([[1.0000, 0.0000, 0.0000, 0.0000]], grad_fn=<CumprodBackward>),\n", |
||
| 3171 | " tensor([[1]]),\n", |
||
| 3172 | " None,\n", |
||
| 3173 | " None)" |
||
| 3174 | ] |
||
| 3175 | }, |
||
| 3176 | "execution_count": 3, |
||
| 3177 | "metadata": {}, |
||
| 3178 | "output_type": "execute_result" |
||
| 3179 | } |
||
| 3180 | ], |
||
| 3181 | "source": [ |
||
| 3182 | "x_path = torch.randint(10, size=(500, 1024)).type(torch.FloatTensor)\n", |
||
| 3183 | "self = MIL_Cluster_FC_surv()\n", |
||
| 3184 | "model.forward(x_path=x_path)" |
||
| 3185 | ] |
||
| 3186 | }, |
||
| 3187 | { |
||
| 3188 | "cell_type": "code", |
||
| 3189 | "execution_count": 7, |
||
| 3190 | "metadata": {}, |
||
| 3191 | "outputs": [], |
||
| 3192 | "source": [ |
||
| 3193 | "import os\n", |
||
| 3194 | "fname = os.path.join('/media/ssd1/pan-cancer/tcga_gbm_20x_features/h5_files/TCGA-02-0001-01Z-00-DX1.83fce43e-42ac-4dcd-b156-2908e75f2e47.h5')" |
||
| 3195 | ] |
||
| 3196 | }, |
||
| 3197 | { |
||
| 3198 | "cell_type": "code", |
||
| 3199 | "execution_count": 27, |
||
| 3200 | "metadata": {}, |
||
| 3201 | "outputs": [], |
||
| 3202 | "source": [ |
||
| 3203 | "import h5py\n", |
||
| 3204 | "h5 = h5py.File(fname, \"r\")\n", |
||
| 3205 | "coords = np.array(h5['coords'])" |
||
| 3206 | ] |
||
| 3207 | }, |
||
| 3208 | { |
||
| 3209 | "cell_type": "code", |
||
| 3210 | "execution_count": null, |
||
| 3211 | "metadata": {}, |
||
| 3212 | "outputs": [], |
||
| 3213 | "source": [ |
||
| 3214 | "fm" |
||
| 3215 | ] |
||
| 3216 | }, |
||
| 3217 | { |
||
| 3218 | "cell_type": "code", |
||
| 3219 | "execution_count": 17, |
||
| 3220 | "metadata": {}, |
||
| 3221 | "outputs": [ |
||
| 3222 | { |
||
| 3223 | "data": { |
||
| 3224 | "text/plain": [ |
||
| 3225 | "array([43121, 29428])" |
||
| 3226 | ] |
||
| 3227 | }, |
||
| 3228 | "execution_count": 17, |
||
| 3229 | "metadata": {}, |
||
| 3230 | "output_type": "execute_result" |
||
| 3231 | } |
||
| 3232 | ], |
||
| 3233 | "source": [ |
||
| 3234 | "np.array(h5['coords'])[0]" |
||
| 3235 | ] |
||
| 3236 | }, |
||
| 3237 | { |
||
| 3238 | "cell_type": "code", |
||
| 3239 | "execution_count": 19, |
||
| 3240 | "metadata": {}, |
||
| 3241 | "outputs": [ |
||
| 3242 | { |
||
| 3243 | "data": { |
||
| 3244 | "text/plain": [ |
||
| 3245 | "array([43121, 29940])" |
||
| 3246 | ] |
||
| 3247 | }, |
||
| 3248 | "execution_count": 19, |
||
| 3249 | "metadata": {}, |
||
| 3250 | "output_type": "execute_result" |
||
| 3251 | } |
||
| 3252 | ], |
||
| 3253 | "source": [ |
||
| 3254 | "np.array(h5['coords'])[1]" |
||
| 3255 | ] |
||
| 3256 | }, |
||
| 3257 | { |
||
| 3258 | "cell_type": "code", |
||
| 3259 | "execution_count": 20, |
||
| 3260 | "metadata": {}, |
||
| 3261 | "outputs": [ |
||
| 3262 | { |
||
| 3263 | "data": { |
||
| 3264 | "text/plain": [ |
||
| 3265 | "512" |
||
| 3266 | ] |
||
| 3267 | }, |
||
| 3268 | "execution_count": 20, |
||
| 3269 | "metadata": {}, |
||
| 3270 | "output_type": "execute_result" |
||
| 3271 | } |
||
| 3272 | ], |
||
| 3273 | "source": [ |
||
| 3274 | "np.array(h5['coords'])[1][1] - np.array(h5['coords'])[0][1]" |
||
| 3275 | ] |
||
| 3276 | }, |
||
| 3277 | { |
||
| 3278 | "cell_type": "code", |
||
| 3279 | "execution_count": 21, |
||
| 3280 | "metadata": {}, |
||
| 3281 | "outputs": [ |
||
| 3282 | { |
||
| 3283 | "data": { |
||
| 3284 | "text/plain": [ |
||
| 3285 | "512" |
||
| 3286 | ] |
||
| 3287 | }, |
||
| 3288 | "execution_count": 21, |
||
| 3289 | "metadata": {}, |
||
| 3290 | "output_type": "execute_result" |
||
| 3291 | } |
||
| 3292 | ], |
||
| 3293 | "source": [ |
||
| 3294 | "np.array(h5['coords'])[2][1] - np.array(h5['coords'])[1][1]" |
||
| 3295 | ] |
||
| 3296 | }, |
||
| 3297 | { |
||
| 3298 | "cell_type": "code", |
||
| 3299 | "execution_count": 23, |
||
| 3300 | "metadata": {}, |
||
| 3301 | "outputs": [], |
||
| 3302 | "source": [ |
||
| 3303 | "import nmslib\n", |
||
| 3304 | "class Hnsw:\n", |
||
| 3305 | "\n", |
||
| 3306 | " def __init__(self, space='cosinesimil', index_params=None,\n", |
||
| 3307 | " query_params=None, print_progress=True):\n", |
||
| 3308 | " self.space = space\n", |
||
| 3309 | " self.index_params = index_params\n", |
||
| 3310 | " self.query_params = query_params\n", |
||
| 3311 | " self.print_progress = print_progress\n", |
||
| 3312 | "\n", |
||
| 3313 | " def fit(self, X):\n", |
||
| 3314 | " index_params = self.index_params\n", |
||
| 3315 | " if index_params is None:\n", |
||
| 3316 | " index_params = {'M': 16, 'post': 0, 'efConstruction': 400}\n", |
||
| 3317 | "\n", |
||
| 3318 | " query_params = self.query_params\n", |
||
| 3319 | " if query_params is None:\n", |
||
| 3320 | " query_params = {'ef': 90}\n", |
||
| 3321 | "\n", |
||
| 3322 | " # this is the actual nmslib part, hopefully the syntax should\n", |
||
| 3323 | " # be pretty readable, the documentation also has a more verbiage\n", |
||
| 3324 | " # introduction: https://nmslib.github.io/nmslib/quickstart.html\n", |
||
| 3325 | " index = nmslib.init(space=self.space, method='hnsw')\n", |
||
| 3326 | " index.addDataPointBatch(X)\n", |
||
| 3327 | " index.createIndex(index_params, print_progress=self.print_progress)\n", |
||
| 3328 | " index.setQueryTimeParams(query_params)\n", |
||
| 3329 | "\n", |
||
| 3330 | " self.index_ = index\n", |
||
| 3331 | " self.index_params_ = index_params\n", |
||
| 3332 | " self.query_params_ = query_params\n", |
||
| 3333 | " return self\n", |
||
| 3334 | "\n", |
||
| 3335 | " def query(self, vector, topn):\n", |
||
| 3336 | " # the knnQuery returns indices and corresponding distance\n", |
||
| 3337 | " # we will throw the distance away for now\n", |
||
| 3338 | " indices, _ = self.index_.knnQuery(vector, k=topn)\n", |
||
| 3339 | " return indices" |
||
| 3340 | ] |
||
| 3341 | }, |
||
| 3342 | { |
||
| 3343 | "cell_type": "code", |
||
| 3344 | "execution_count": null, |
||
| 3345 | "metadata": {}, |
||
| 3346 | "outputs": [], |
||
| 3347 | "source": [ |
||
| 3348 | "x" |
||
| 3349 | ] |
||
| 3350 | }, |
||
| 3351 | { |
||
| 3352 | "cell_type": "code", |
||
| 3353 | "execution_count": 54, |
||
| 3354 | "metadata": {}, |
||
| 3355 | "outputs": [ |
||
| 3356 | { |
||
| 3357 | "data": { |
||
| 3358 | "text/plain": [ |
||
| 3359 | "array([85, 87, 88, 73, 75, 76, 63, 29], dtype=int32)" |
||
| 3360 | ] |
||
| 3361 | }, |
||
| 3362 | "execution_count": 54, |
||
| 3363 | "metadata": {}, |
||
| 3364 | "output_type": "execute_result" |
||
| 3365 | } |
||
| 3366 | ], |
||
| 3367 | "source": [ |
||
| 3368 | "model = Hnsw(space='l2')\n", |
||
| 3369 | "model.fit(coords)\n", |
||
| 3370 | "model.query(coords, topn=8)" |
||
| 3371 | ] |
||
| 3372 | }, |
||
| 3373 | { |
||
| 3374 | "cell_type": "code", |
||
| 3375 | "execution_count": 59, |
||
| 3376 | "metadata": {}, |
||
| 3377 | "outputs": [], |
||
| 3378 | "source": [ |
||
| 3379 | "import networkx as nx\n", |
||
| 3380 | "G = nx.Graph()\n" |
||
| 3381 | ] |
||
| 3382 | }, |
||
| 3383 | { |
||
| 3384 | "cell_type": "code", |
||
| 3385 | "execution_count": 56, |
||
| 3386 | "metadata": {}, |
||
| 3387 | "outputs": [ |
||
| 3388 | { |
||
| 3389 | "data": { |
||
| 3390 | "text/plain": [ |
||
| 3391 | "array([43121, 29428])" |
||
| 3392 | ] |
||
| 3393 | }, |
||
| 3394 | "execution_count": 56, |
||
| 3395 | "metadata": {}, |
||
| 3396 | "output_type": "execute_result" |
||
| 3397 | } |
||
| 3398 | ], |
||
| 3399 | "source": [ |
||
| 3400 | "for" |
||
| 3401 | ] |
||
| 3402 | }, |
||
| 3403 | { |
||
| 3404 | "cell_type": "code", |
||
| 3405 | "execution_count": 52, |
||
| 3406 | "metadata": {}, |
||
| 3407 | "outputs": [ |
||
| 3408 | { |
||
| 3409 | "data": { |
||
| 3410 | "text/plain": [ |
||
| 3411 | "130" |
||
| 3412 | ] |
||
| 3413 | }, |
||
| 3414 | "execution_count": 52, |
||
| 3415 | "metadata": {}, |
||
| 3416 | "output_type": "execute_result" |
||
| 3417 | } |
||
| 3418 | ], |
||
| 3419 | "source": [ |
||
| 3420 | "temp[3]" |
||
| 3421 | ] |
||
| 3422 | }, |
||
| 3423 | { |
||
| 3424 | "cell_type": "code", |
||
| 3425 | "execution_count": null, |
||
| 3426 | "metadata": {}, |
||
| 3427 | "outputs": [], |
||
| 3428 | "source": [ |
||
| 3429 | "model" |
||
| 3430 | ] |
||
| 3431 | }, |
||
| 3432 | { |
||
| 3433 | "cell_type": "code", |
||
| 3434 | "execution_count": 29, |
||
| 3435 | "metadata": {}, |
||
| 3436 | "outputs": [ |
||
| 3437 | { |
||
| 3438 | "data": { |
||
| 3439 | "text/plain": [ |
||
| 3440 | "array([ 7440, 13280])" |
||
| 3441 | ] |
||
| 3442 | }, |
||
| 3443 | "execution_count": 29, |
||
| 3444 | "metadata": {}, |
||
| 3445 | "output_type": "execute_result" |
||
| 3446 | } |
||
| 3447 | ], |
||
| 3448 | "source": [ |
||
| 3449 | "coords[100]" |
||
| 3450 | ] |
||
| 3451 | }, |
||
| 3452 | { |
||
| 3453 | "cell_type": "code", |
||
| 3454 | "execution_count": 33, |
||
| 3455 | "metadata": {}, |
||
| 3456 | "outputs": [], |
||
| 3457 | "source": [ |
||
| 3458 | "indices = model.query(coords[100], topn =10)" |
||
| 3459 | ] |
||
| 3460 | }, |
||
| 3461 | { |
||
| 3462 | "cell_type": "code", |
||
| 3463 | "execution_count": 34, |
||
| 3464 | "metadata": {}, |
||
| 3465 | "outputs": [ |
||
| 3466 | { |
||
| 3467 | "data": { |
||
| 3468 | "text/plain": [ |
||
| 3469 | "array([[ 7440, 13280],\n", |
||
| 3470 | " [ 7440, 13792],\n", |
||
| 3471 | " [ 7952, 13280],\n", |
||
| 3472 | " [ 6928, 13792],\n", |
||
| 3473 | " [ 7952, 12768],\n", |
||
| 3474 | " [ 7952, 13792],\n", |
||
| 3475 | " [ 7440, 14304],\n", |
||
| 3476 | " [ 8464, 13280],\n", |
||
| 3477 | " [ 6928, 14304],\n", |
||
| 3478 | " [ 8464, 13792]])" |
||
| 3479 | ] |
||
| 3480 | }, |
||
| 3481 | "execution_count": 34, |
||
| 3482 | "metadata": {}, |
||
| 3483 | "output_type": "execute_result" |
||
| 3484 | } |
||
| 3485 | ], |
||
| 3486 | "source": [ |
||
| 3487 | "coords[indices]" |
||
| 3488 | ] |
||
| 3489 | }, |
||
| 3490 | { |
||
| 3491 | "cell_type": "code", |
||
| 3492 | "execution_count": 84, |
||
| 3493 | "metadata": {}, |
||
| 3494 | "outputs": [], |
||
| 3495 | "source": [ |
||
| 3496 | "def do_KmeansPCA(X=None, y=None, scaler=None, n_clusters=4, n_components=5):\n", |
||
| 3497 | " import pandas as pd\n", |
||
| 3498 | " import seaborn as sns\n", |
||
| 3499 | " from sklearn.datasets import make_blobs\n", |
||
| 3500 | " from sklearn import decomposition\n", |
||
| 3501 | " from sklearn.decomposition import PCA, TruncatedSVD\n", |
||
| 3502 | " from sklearn.preprocessing import StandardScaler, Normalizer\n", |
||
| 3503 | " from sklearn.pipeline import make_pipeline\n", |
||
| 3504 | " from sklearn.cluster import KMeans\n", |
||
| 3505 | " ### Initialize Scaler\n", |
||
| 3506 | " if scaler is None: \n", |
||
| 3507 | " scaler = StandardScaler()\n", |
||
| 3508 | " ### Get Random Data\n", |
||
| 3509 | " X, y = make_blobs(n_features=10, n_samples=100, centers=4, random_state=4, cluster_std=7)\n", |
||
| 3510 | " ### Scale Data\n", |
||
| 3511 | " X = scaler.fit_transform(X)\n", |
||
| 3512 | " ### Perform K-Means Clustering\n", |
||
| 3513 | " cls = KMeans(n_clusters=n_clusters, init='k-means++', n_jobs=-1, n_init=1)\n", |
||
| 3514 | " y_pred = cls.fit_predict(X)\n", |
||
| 3515 | " ### Perform PCA\n", |
||
| 3516 | " pca = PCA(n_components=n_components)\n", |
||
| 3517 | " pc = pca.fit_transform(X)\n", |
||
| 3518 | " ### Plot Results\n", |
||
| 3519 | " columns = ['PC%d'%c for c in range(1, n_components+1)]\n", |
||
| 3520 | " pc_df = pd.DataFrame(data=pc, columns=columns)\n", |
||
| 3521 | " pc_df['y_pred'] = y_pred\n", |
||
| 3522 | " pc_df['y'] = y\n", |
||
| 3523 | " df = pd.DataFrame({'Variance Explained':pca.explained_variance_ratio_, 'Principal Components': columns})\n", |
||
| 3524 | " sns.barplot(x='Principal Components',y=\"Variance Explained\", data=df, color=\"c\")\n", |
||
| 3525 | " sns.lmplot( x=\"PC1\", y=\"PC2\", data=pc_df, fit_reg=False, \n", |
||
| 3526 | " hue='y', legend=True, scatter_kws={\"s\": 80})\n", |
||
| 3527 | " sns.lmplot( x=\"PC1\", y=\"PC2\", data=pc_df, fit_reg=False, \n", |
||
| 3528 | " hue='y', legend=True, scatter_kws={\"s\": 80})" |
||
| 3529 | ] |
||
| 3530 | }, |
||
| 3531 | { |
||
| 3532 | "cell_type": "code", |
||
| 3533 | "execution_count": 85, |
||
| 3534 | "metadata": {}, |
||
| 3535 | "outputs": [ |
||
| 3536 | { |
||
| 3537 | "data": { |
||
| 3538 | "image/png": "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\n", |
||
| 3539 | "text/plain": [ |
||
| 3540 | "<Figure size 432x288 with 1 Axes>" |
||
| 3541 | ] |
||
| 3542 | }, |
||
| 3543 | "metadata": { |
||
| 3544 | "needs_background": "light" |
||
| 3545 | }, |
||
| 3546 | "output_type": "display_data" |
||
| 3547 | }, |
||
| 3548 | { |
||
| 3549 | "data": { |
||
| 3550 | "image/png": "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\n", |
||
| 3551 | "text/plain": [ |
||
| 3552 | "<Figure size 402.375x360 with 1 Axes>" |
||
| 3553 | ] |
||
| 3554 | }, |
||
| 3555 | "metadata": { |
||
| 3556 | "needs_background": "light" |
||
| 3557 | }, |
||
| 3558 | "output_type": "display_data" |
||
| 3559 | }, |
||
| 3560 | { |
||
| 3561 | "data": { |
||
| 3562 | "image/png": "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\n", |
||
| 3563 | "text/plain": [ |
||
| 3564 | "<Figure size 402.375x360 with 1 Axes>" |
||
| 3565 | ] |
||
| 3566 | }, |
||
| 3567 | "metadata": { |
||
| 3568 | "needs_background": "light" |
||
| 3569 | }, |
||
| 3570 | "output_type": "display_data" |
||
| 3571 | } |
||
| 3572 | ], |
||
| 3573 | "source": [ |
||
| 3574 | "do_KmeansPCA()" |
||
| 3575 | ] |
||
| 3576 | }, |
||
| 3577 | { |
||
| 3578 | "cell_type": "code", |
||
| 3579 | "execution_count": 76, |
||
| 3580 | "metadata": {}, |
||
| 3581 | "outputs": [ |
||
| 3582 | { |
||
| 3583 | "data": { |
||
| 3584 | "text/plain": [ |
||
| 3585 | "(tensor([[1.0000, 0.0000, 1.0000, 0.9998]], grad_fn=<SigmoidBackward>),\n", |
||
| 3586 | " tensor([[0., 0., 0., 0.]], grad_fn=<CumprodBackward>),\n", |
||
| 3587 | " tensor([[2]]),\n", |
||
| 3588 | " None,\n", |
||
| 3589 | " None)" |
||
| 3590 | ] |
||
| 3591 | }, |
||
| 3592 | "execution_count": 76, |
||
| 3593 | "metadata": {}, |
||
| 3594 | "output_type": "execute_result" |
||
| 3595 | } |
||
| 3596 | ], |
||
| 3597 | "source": [ |
||
| 3598 | "model.forward(x_path=x_path)" |
||
| 3599 | ] |
||
| 3600 | }, |
||
| 3601 | { |
||
| 3602 | "cell_type": "code", |
||
| 3603 | "execution_count": 69, |
||
| 3604 | "metadata": {}, |
||
| 3605 | "outputs": [ |
||
| 3606 | { |
||
| 3607 | "ename": "SyntaxError", |
||
| 3608 | "evalue": "invalid syntax (<ipython-input-69-c543913fa78f>, line 1)", |
||
| 3609 | "output_type": "error", |
||
| 3610 | "traceback": [ |
||
| 3611 | "\u001b[0;36m File \u001b[0;32m\"<ipython-input-69-c543913fa78f>\"\u001b[0;36m, line \u001b[0;32m1\u001b[0m\n\u001b[0;31m import ..models\u001b[0m\n\u001b[0m ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m invalid syntax\n" |
||
| 3612 | ] |
||
| 3613 | } |
||
| 3614 | ], |
||
| 3615 | "source": [ |
||
| 3616 | "import ..models" |
||
| 3617 | ] |
||
| 3618 | }, |
||
| 3619 | { |
||
| 3620 | "cell_type": "code", |
||
| 3621 | "execution_count": 63, |
||
| 3622 | "metadata": {}, |
||
| 3623 | "outputs": [], |
||
| 3624 | "source": [ |
||
| 3625 | "x_path = torch.randint(10, size=(500, 1024)).type(torch.FloatTensor)" |
||
| 3626 | ] |
||
| 3627 | }, |
||
| 3628 | { |
||
| 3629 | "cell_type": "code", |
||
| 3630 | "execution_count": 65, |
||
| 3631 | "metadata": {}, |
||
| 3632 | "outputs": [ |
||
| 3633 | { |
||
| 3634 | "ename": "NameError", |
||
| 3635 | "evalue": "name 'MultiheadAttention' is not defined", |
||
| 3636 | "output_type": "error", |
||
| 3637 | "traceback": [ |
||
| 3638 | "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", |
||
| 3639 | "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", |
||
| 3640 | "\u001b[0;32m<ipython-input-65-f85a99af33ee>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mself\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMM_CoAttn_Transformer_Surv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0momic_sizes\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msig_sizes\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m", |
||
| 3641 | "\u001b[0;32m<ipython-input-62-9e5f322e30a0>\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, omic_sizes, n_classes, model_size_wsi, model_size_omic, dropout)\u001b[0m\n\u001b[1;32m 28\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 29\u001b[0m \u001b[0;31m### Multihead Attention\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcoattn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMultiheadAttention\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0membed_dim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m256\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_heads\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 31\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 32\u001b[0m \u001b[0;31m### Transformer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", |
||
| 3642 | "\u001b[0;31mNameError\u001b[0m: name 'MultiheadAttention' is not defined" |
||
| 3643 | ] |
||
| 3644 | } |
||
| 3645 | ], |
||
| 3646 | "source": [ |
||
| 3647 | "self = MM_CoAttn_Transformer_Surv(omic_sizes=sig_sizes)" |
||
| 3648 | ] |
||
| 3649 | }, |
||
| 3650 | { |
||
| 3651 | "cell_type": "code", |
||
| 3652 | "execution_count": 52, |
||
| 3653 | "metadata": {}, |
||
| 3654 | "outputs": [ |
||
| 3655 | { |
||
| 3656 | "ename": "NameError", |
||
| 3657 | "evalue": "name 'sig_size' is not defined", |
||
| 3658 | "output_type": "error", |
||
| 3659 | "traceback": [ |
||
| 3660 | "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", |
||
| 3661 | "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", |
||
| 3662 | "\u001b[0;32m<ipython-input-52-097a03ed0c40>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mself\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMM_CoAttn_Surv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msig_sizes\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msig_sizes\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0mx_path\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m500\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1024\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mFloatTensor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0msig_feats\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mFloatTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0msize\u001b[0m \u001b[0;32min\u001b[0m \u001b[0msig_sizes\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0mx_path\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mattention_net\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munsqueeze\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", |
||
| 3663 | "\u001b[0;32m<ipython-input-43-4469ba9e1eea>\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, sig_sizes, n_classes, model_size_wsi, model_size_omic, dropout)\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[0mhidden\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize_dict_omic\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mmodel_size_omic\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 20\u001b[0m \u001b[0msig_networks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m \u001b[0;32mfor\u001b[0m \u001b[0minput_dim\u001b[0m \u001b[0;32min\u001b[0m \u001b[0msig_size\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 22\u001b[0m \u001b[0mfc_omic\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mSNN_Block\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdim1\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minput_dim\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdim2\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mhidden\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 23\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mhidden\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", |
||
| 3664 | "\u001b[0;31mNameError\u001b[0m: name 'sig_size' is not defined" |
||
| 3665 | ] |
||
| 3666 | } |
||
| 3667 | ], |
||
| 3668 | "source": [ |
||
| 3669 | "self = MM_CoAttn_Surv(sig_sizes=sig_sizes)\n", |
||
| 3670 | "x_path = torch.randint(10, size=(500, 1024)).type(torch.FloatTensor)\n", |
||
| 3671 | "sig_feats = [torch.randint(10, size=(size,)).type(torch.FloatTensor) for size in sig_sizes]\n", |
||
| 3672 | "\n", |
||
| 3673 | "x_path = self.attention_net(x_path).unsqueeze(1)\n", |
||
| 3674 | "x_omic = torch.stack([self.sig_networks[idx].forward(sig_feat) for idx, sig_feat in enumerate(sig_feats)]).unsqueeze(1)\n", |
||
| 3675 | "\n", |
||
| 3676 | "out, attention_weights = self.coattn(x_omic, x_path, x_path)\n", |
||
| 3677 | "out = self.transformer(out)\n", |
||
| 3678 | "out = self.conv(out.squeeze(1).T.unsqueeze(0))\n", |
||
| 3679 | "#out = self.classifier(out.squeeze(0).squeeze(1))" |
||
| 3680 | ] |
||
| 3681 | }, |
||
| 3682 | { |
||
| 3683 | "cell_type": "code", |
||
| 3684 | "execution_count": 471, |
||
| 3685 | "metadata": {}, |
||
| 3686 | "outputs": [ |
||
| 3687 | { |
||
| 3688 | "data": { |
||
| 3689 | "text/plain": [ |
||
| 3690 | "torch.Size([1, 256, 1])" |
||
| 3691 | ] |
||
| 3692 | }, |
||
| 3693 | "execution_count": 471, |
||
| 3694 | "metadata": {}, |
||
| 3695 | "output_type": "execute_result" |
||
| 3696 | } |
||
| 3697 | ], |
||
| 3698 | "source": [ |
||
| 3699 | "out.shape" |
||
| 3700 | ] |
||
| 3701 | }, |
||
| 3702 | { |
||
| 3703 | "cell_type": "code", |
||
| 3704 | "execution_count": 472, |
||
| 3705 | "metadata": {}, |
||
| 3706 | "outputs": [ |
||
| 3707 | { |
||
| 3708 | "data": { |
||
| 3709 | "text/plain": [ |
||
| 3710 | "tensor([[[ 0.5998, 1.9873, -1.1435, ..., -0.0048, 0.2963, 1.1112]],\n", |
||
| 3711 | "\n", |
||
| 3712 | " [[-0.4201, -0.1456, 0.2057, ..., -0.2175, 0.4188, 0.4702]],\n", |
||
| 3713 | "\n", |
||
| 3714 | " [[ 1.0294, 3.1634, 0.4595, ..., 1.2059, 0.5845, 1.4114]],\n", |
||
| 3715 | "\n", |
||
| 3716 | " [[-1.1435, -1.1435, -1.1435, ..., 0.1951, -0.4378, 0.2051]],\n", |
||
| 3717 | "\n", |
||
| 3718 | " [[ 0.9948, 1.1596, 2.1419, ..., -0.1225, 1.3597, -0.3037]],\n", |
||
| 3719 | "\n", |
||
| 3720 | " [[ 0.4019, -1.1435, -0.1522, ..., -0.2058, 0.0351, -1.1435]]],\n", |
||
| 3721 | " grad_fn=<UnsqueezeBackward0>)" |
||
| 3722 | ] |
||
| 3723 | }, |
||
| 3724 | "execution_count": 472, |
||
| 3725 | "metadata": {}, |
||
| 3726 | "output_type": "execute_result" |
||
| 3727 | } |
||
| 3728 | ], |
||
| 3729 | "source": [ |
||
| 3730 | "x_omic" |
||
| 3731 | ] |
||
| 3732 | }, |
||
| 3733 | { |
||
| 3734 | "cell_type": "code", |
||
| 3735 | "execution_count": 474, |
||
| 3736 | "metadata": {}, |
||
| 3737 | "outputs": [], |
||
| 3738 | "source": [ |
||
| 3739 | "self = MM_CoAttn_Surv(sig_sizes=sig_sizes)\n", |
||
| 3740 | "x_path = torch.randint(10, size=(500, 1024)).type(torch.FloatTensor)\n", |
||
| 3741 | "sig_feats = [torch.randint(10, size=(size,)).type(torch.FloatTensor) for size in sig_sizes]\n", |
||
| 3742 | "\n", |
||
| 3743 | "x_path = self.attention_net(x_path).unsqueeze(1)\n", |
||
| 3744 | "x_omic = torch.stack([self.sig_networks[idx].forward(sig_feat) for idx, sig_feat in enumerate(sig_feats)]).unsqueeze(1)\n", |
||
| 3745 | "out, attention_weights = self.coattn(x_omic, x_path, x_path)\n", |
||
| 3746 | "\n", |
||
| 3747 | "out = self.transformer(out)\n" |
||
| 3748 | ] |
||
| 3749 | }, |
||
| 3750 | { |
||
| 3751 | "cell_type": "code", |
||
| 3752 | "execution_count": 491, |
||
| 3753 | "metadata": {}, |
||
| 3754 | "outputs": [ |
||
| 3755 | { |
||
| 3756 | "data": { |
||
| 3757 | "text/plain": [ |
||
| 3758 | "torch.Size([1536])" |
||
| 3759 | ] |
||
| 3760 | }, |
||
| 3761 | "execution_count": 491, |
||
| 3762 | "metadata": {}, |
||
| 3763 | "output_type": "execute_result" |
||
| 3764 | } |
||
| 3765 | ], |
||
| 3766 | "source": [ |
||
| 3767 | "torch.cat([self.sig_networks[idx].forward(sig_feat) for idx, sig_feat in enumerate(sig_feats)]).shape" |
||
| 3768 | ] |
||
| 3769 | }, |
||
| 3770 | { |
||
| 3771 | "cell_type": "code", |
||
| 3772 | "execution_count": 484, |
||
| 3773 | "metadata": {}, |
||
| 3774 | "outputs": [ |
||
| 3775 | { |
||
| 3776 | "data": { |
||
| 3777 | "text/plain": [ |
||
| 3778 | "torch.Size([6, 1, 512])" |
||
| 3779 | ] |
||
| 3780 | }, |
||
| 3781 | "execution_count": 484, |
||
| 3782 | "metadata": {}, |
||
| 3783 | "output_type": "execute_result" |
||
| 3784 | } |
||
| 3785 | ], |
||
| 3786 | "source": [ |
||
| 3787 | "torch.cat([out, out], axis=2).shape" |
||
| 3788 | ] |
||
| 3789 | }, |
||
| 3790 | { |
||
| 3791 | "cell_type": "code", |
||
| 3792 | "execution_count": 455, |
||
| 3793 | "metadata": {}, |
||
| 3794 | "outputs": [ |
||
| 3795 | { |
||
| 3796 | "data": { |
||
| 3797 | "text/plain": [ |
||
| 3798 | "torch.Size([6, 1, 256])" |
||
| 3799 | ] |
||
| 3800 | }, |
||
| 3801 | "execution_count": 455, |
||
| 3802 | "metadata": {}, |
||
| 3803 | "output_type": "execute_result" |
||
| 3804 | } |
||
| 3805 | ], |
||
| 3806 | "source": [ |
||
| 3807 | "out.shape" |
||
| 3808 | ] |
||
| 3809 | }, |
||
| 3810 | { |
||
| 3811 | "cell_type": "code", |
||
| 3812 | "execution_count": 452, |
||
| 3813 | "metadata": {}, |
||
| 3814 | "outputs": [ |
||
| 3815 | { |
||
| 3816 | "data": { |
||
| 3817 | "text/plain": [ |
||
| 3818 | "torch.Size([6, 1, 256])" |
||
| 3819 | ] |
||
| 3820 | }, |
||
| 3821 | "execution_count": 452, |
||
| 3822 | "metadata": {}, |
||
| 3823 | "output_type": "execute_result" |
||
| 3824 | } |
||
| 3825 | ], |
||
| 3826 | "source": [ |
||
| 3827 | "out.shape" |
||
| 3828 | ] |
||
| 3829 | }, |
||
| 3830 | { |
||
| 3831 | "cell_type": "code", |
||
| 3832 | "execution_count": null, |
||
| 3833 | "metadata": {}, |
||
| 3834 | "outputs": [], |
||
| 3835 | "source": [] |
||
| 3836 | }, |
||
| 3837 | { |
||
| 3838 | "cell_type": "code", |
||
| 3839 | "execution_count": 423, |
||
| 3840 | "metadata": {}, |
||
| 3841 | "outputs": [ |
||
| 3842 | { |
||
| 3843 | "data": { |
||
| 3844 | "text/plain": [ |
||
| 3845 | "torch.Size([1, 8, 6, 500])" |
||
| 3846 | ] |
||
| 3847 | }, |
||
| 3848 | "execution_count": 423, |
||
| 3849 | "metadata": {}, |
||
| 3850 | "output_type": "execute_result" |
||
| 3851 | } |
||
| 3852 | ], |
||
| 3853 | "source": [ |
||
| 3854 | "attention_weights.shape" |
||
| 3855 | ] |
||
| 3856 | }, |
||
| 3857 | { |
||
| 3858 | "cell_type": "code", |
||
| 3859 | "execution_count": 415, |
||
| 3860 | "metadata": {}, |
||
| 3861 | "outputs": [ |
||
| 3862 | { |
||
| 3863 | "data": { |
||
| 3864 | "text/plain": [ |
||
| 3865 | "tensor([[[0.0018, 0.0020, 0.0012, ..., 0.0016, 0.0025, 0.0031],\n", |
||
| 3866 | " [0.0026, 0.0015, 0.0016, ..., 0.0021, 0.0021, 0.0016],\n", |
||
| 3867 | " [0.0019, 0.0014, 0.0011, ..., 0.0020, 0.0013, 0.0025],\n", |
||
| 3868 | " [0.0016, 0.0013, 0.0023, ..., 0.0009, 0.0015, 0.0027],\n", |
||
| 3869 | " [0.0015, 0.0013, 0.0023, ..., 0.0026, 0.0019, 0.0026],\n", |
||
| 3870 | " [0.0013, 0.0019, 0.0025, ..., 0.0022, 0.0020, 0.0021]]],\n", |
||
| 3871 | " grad_fn=<DivBackward0>)" |
||
| 3872 | ] |
||
| 3873 | }, |
||
| 3874 | "execution_count": 415, |
||
| 3875 | "metadata": {}, |
||
| 3876 | "output_type": "execute_result" |
||
| 3877 | } |
||
| 3878 | ], |
||
| 3879 | "source": [ |
||
| 3880 | "attention_weights_0" |
||
| 3881 | ] |
||
| 3882 | }, |
||
| 3883 | { |
||
| 3884 | "cell_type": "code", |
||
| 3885 | "execution_count": 416, |
||
| 3886 | "metadata": {}, |
||
| 3887 | "outputs": [ |
||
| 3888 | { |
||
| 3889 | "data": { |
||
| 3890 | "text/plain": [ |
||
| 3891 | "tensor([[[0.0018, 0.0020, 0.0012, ..., 0.0016, 0.0025, 0.0031],\n", |
||
| 3892 | " [0.0026, 0.0015, 0.0016, ..., 0.0021, 0.0021, 0.0016],\n", |
||
| 3893 | " [0.0019, 0.0014, 0.0011, ..., 0.0020, 0.0013, 0.0025],\n", |
||
| 3894 | " [0.0016, 0.0013, 0.0023, ..., 0.0009, 0.0015, 0.0027],\n", |
||
| 3895 | " [0.0015, 0.0013, 0.0023, ..., 0.0026, 0.0019, 0.0026],\n", |
||
| 3896 | " [0.0013, 0.0019, 0.0025, ..., 0.0022, 0.0020, 0.0021]]],\n", |
||
| 3897 | " grad_fn=<DivBackward0>)" |
||
| 3898 | ] |
||
| 3899 | }, |
||
| 3900 | "execution_count": 416, |
||
| 3901 | "metadata": {}, |
||
| 3902 | "output_type": "execute_result" |
||
| 3903 | } |
||
| 3904 | ], |
||
| 3905 | "source": [ |
||
| 3906 | "softmax(attention_weights_1, dim=-1).sum(axis=1) / 8" |
||
| 3907 | ] |
||
| 3908 | }, |
||
| 3909 | { |
||
| 3910 | "cell_type": "code", |
||
| 3911 | "execution_count": 411, |
||
| 3912 | "metadata": {}, |
||
| 3913 | "outputs": [ |
||
| 3914 | { |
||
| 3915 | "data": { |
||
| 3916 | "text/plain": [ |
||
| 3917 | "torch.Size([1, 1, 6, 500])" |
||
| 3918 | ] |
||
| 3919 | }, |
||
| 3920 | "execution_count": 411, |
||
| 3921 | "metadata": {}, |
||
| 3922 | "output_type": "execute_result" |
||
| 3923 | } |
||
| 3924 | ], |
||
| 3925 | "source": [ |
||
| 3926 | "softmax(attention_weights_1, dim=-1).shape" |
||
| 3927 | ] |
||
| 3928 | }, |
||
| 3929 | { |
||
| 3930 | "cell_type": "code", |
||
| 3931 | "execution_count": 339, |
||
| 3932 | "metadata": {}, |
||
| 3933 | "outputs": [ |
||
| 3934 | { |
||
| 3935 | "data": { |
||
| 3936 | "text/plain": [ |
||
| 3937 | "tensor(1.0000, grad_fn=<SumBackward0>)" |
||
| 3938 | ] |
||
| 3939 | }, |
||
| 3940 | "execution_count": 339, |
||
| 3941 | "metadata": {}, |
||
| 3942 | "output_type": "execute_result" |
||
| 3943 | } |
||
| 3944 | ], |
||
| 3945 | "source": [ |
||
| 3946 | "attention_weights_0[0][0].sum()" |
||
| 3947 | ] |
||
| 3948 | }, |
||
| 3949 | { |
||
| 3950 | "cell_type": "code", |
||
| 3951 | "execution_count": 396, |
||
| 3952 | "metadata": {}, |
||
| 3953 | "outputs": [], |
||
| 3954 | "source": [ |
||
| 3955 | "test = softmax(attention_weights_2, dim=-1)" |
||
| 3956 | ] |
||
| 3957 | }, |
||
| 3958 | { |
||
| 3959 | "cell_type": "code", |
||
| 3960 | "execution_count": 402, |
||
| 3961 | "metadata": {}, |
||
| 3962 | "outputs": [ |
||
| 3963 | { |
||
| 3964 | "data": { |
||
| 3965 | "text/plain": [ |
||
| 3966 | "tensor([0.0024, 0.0030, 0.0019, 0.0018, 0.0038, 0.0015, 0.0020, 0.0016, 0.0015,\n", |
||
| 3967 | " 0.0019, 0.0015, 0.0035, 0.0026, 0.0017, 0.0014, 0.0013, 0.0023, 0.0020,\n", |
||
| 3968 | " 0.0017, 0.0010], grad_fn=<SliceBackward>)" |
||
| 3969 | ] |
||
| 3970 | }, |
||
| 3971 | "execution_count": 402, |
||
| 3972 | "metadata": {}, |
||
| 3973 | "output_type": "execute_result" |
||
| 3974 | } |
||
| 3975 | ], |
||
| 3976 | "source": [ |
||
| 3977 | "attention_weights_0[0][0][:20]" |
||
| 3978 | ] |
||
| 3979 | }, |
||
| 3980 | { |
||
| 3981 | "cell_type": "code", |
||
| 3982 | "execution_count": 404, |
||
| 3983 | "metadata": {}, |
||
| 3984 | "outputs": [ |
||
| 3985 | { |
||
| 3986 | "data": { |
||
| 3987 | "text/plain": [ |
||
| 3988 | "tensor([0.0028, 0.0033, 0.0019, 0.0013, 0.0042, 0.0016, 0.0024, 0.0018, 0.0019,\n", |
||
| 3989 | " 0.0024, 0.0016, 0.0033, 0.0022, 0.0014, 0.0016, 0.0013, 0.0023, 0.0021,\n", |
||
| 3990 | " 0.0013, 0.0013], grad_fn=<SliceBackward>)" |
||
| 3991 | ] |
||
| 3992 | }, |
||
| 3993 | "execution_count": 404, |
||
| 3994 | "metadata": {}, |
||
| 3995 | "output_type": "execute_result" |
||
| 3996 | } |
||
| 3997 | ], |
||
| 3998 | "source": [ |
||
| 3999 | "test[0][0][:20]" |
||
| 4000 | ] |
||
| 4001 | }, |
||
| 4002 | { |
||
| 4003 | "cell_type": "code", |
||
| 4004 | "execution_count": 366, |
||
| 4005 | "metadata": {}, |
||
| 4006 | "outputs": [ |
||
| 4007 | { |
||
| 4008 | "data": { |
||
| 4009 | "text/plain": [ |
||
| 4010 | "tensor([[[False, False, False, ..., False, False, False],\n", |
||
| 4011 | " [False, False, False, ..., False, False, False],\n", |
||
| 4012 | " [False, False, False, ..., False, False, False],\n", |
||
| 4013 | " [False, False, False, ..., False, False, False],\n", |
||
| 4014 | " [False, False, False, ..., False, False, False],\n", |
||
| 4015 | " [False, False, False, ..., False, False, False]]])" |
||
| 4016 | ] |
||
| 4017 | }, |
||
| 4018 | "execution_count": 366, |
||
| 4019 | "metadata": {}, |
||
| 4020 | "output_type": "execute_result" |
||
| 4021 | } |
||
| 4022 | ], |
||
| 4023 | "source": [ |
||
| 4024 | "torch.eq(attention_weights_0, test)" |
||
| 4025 | ] |
||
| 4026 | }, |
||
| 4027 | { |
||
| 4028 | "cell_type": "code", |
||
| 4029 | "execution_count": 320, |
||
| 4030 | "metadata": {}, |
||
| 4031 | "outputs": [ |
||
| 4032 | { |
||
| 4033 | "data": { |
||
| 4034 | "text/plain": [ |
||
| 4035 | "torch.Size([1, 8, 6, 500])" |
||
| 4036 | ] |
||
| 4037 | }, |
||
| 4038 | "execution_count": 320, |
||
| 4039 | "metadata": {}, |
||
| 4040 | "output_type": "execute_result" |
||
| 4041 | } |
||
| 4042 | ], |
||
| 4043 | "source": [ |
||
| 4044 | "attention_weights_1.shape" |
||
| 4045 | ] |
||
| 4046 | }, |
||
| 4047 | { |
||
| 4048 | "cell_type": "code", |
||
| 4049 | "execution_count": 318, |
||
| 4050 | "metadata": {}, |
||
| 4051 | "outputs": [ |
||
| 4052 | { |
||
| 4053 | "data": { |
||
| 4054 | "text/plain": [ |
||
| 4055 | "torch.Size([1, 6, 500])" |
||
| 4056 | ] |
||
| 4057 | }, |
||
| 4058 | "execution_count": 318, |
||
| 4059 | "metadata": {}, |
||
| 4060 | "output_type": "execute_result" |
||
| 4061 | } |
||
| 4062 | ], |
||
| 4063 | "source": [ |
||
| 4064 | "attention_weights_2.shape" |
||
| 4065 | ] |
||
| 4066 | }, |
||
| 4067 | { |
||
| 4068 | "cell_type": "code", |
||
| 4069 | "execution_count": 282, |
||
| 4070 | "metadata": {}, |
||
| 4071 | "outputs": [], |
||
| 4072 | "source": [ |
||
| 4073 | "out = self.classifier(out.squeeze(0).squeeze(1))" |
||
| 4074 | ] |
||
| 4075 | }, |
||
| 4076 | { |
||
| 4077 | "cell_type": "code", |
||
| 4078 | "execution_count": 284, |
||
| 4079 | "metadata": {}, |
||
| 4080 | "outputs": [ |
||
| 4081 | { |
||
| 4082 | "data": { |
||
| 4083 | "text/plain": [ |
||
| 4084 | "tensor([ 0.2832, 0.1548, -0.0972, -0.2801], grad_fn=<AddBackward0>)" |
||
| 4085 | ] |
||
| 4086 | }, |
||
| 4087 | "execution_count": 284, |
||
| 4088 | "metadata": {}, |
||
| 4089 | "output_type": "execute_result" |
||
| 4090 | } |
||
| 4091 | ], |
||
| 4092 | "source": [ |
||
| 4093 | "out" |
||
| 4094 | ] |
||
| 4095 | }, |
||
| 4096 | { |
||
| 4097 | "cell_type": "code", |
||
| 4098 | "execution_count": 269, |
||
| 4099 | "metadata": {}, |
||
| 4100 | "outputs": [ |
||
| 4101 | { |
||
| 4102 | "data": { |
||
| 4103 | "text/plain": [ |
||
| 4104 | "tensor([[0.0018, 0.0019, 0.0019, ..., 0.0019, 0.0022, 0.0018],\n", |
||
| 4105 | " [0.0020, 0.0020, 0.0021, ..., 0.0021, 0.0020, 0.0020],\n", |
||
| 4106 | " [0.0019, 0.0022, 0.0021, ..., 0.0019, 0.0019, 0.0020],\n", |
||
| 4107 | " [0.0021, 0.0022, 0.0019, ..., 0.0018, 0.0020, 0.0021],\n", |
||
| 4108 | " [0.0019, 0.0019, 0.0020, ..., 0.0020, 0.0018, 0.0019],\n", |
||
| 4109 | " [0.0021, 0.0021, 0.0019, ..., 0.0019, 0.0021, 0.0021]],\n", |
||
| 4110 | " grad_fn=<SelectBackward>)" |
||
| 4111 | ] |
||
| 4112 | }, |
||
| 4113 | "execution_count": 269, |
||
| 4114 | "metadata": {}, |
||
| 4115 | "output_type": "execute_result" |
||
| 4116 | } |
||
| 4117 | ], |
||
| 4118 | "source": [ |
||
| 4119 | "attention_weights[0]" |
||
| 4120 | ] |
||
| 4121 | }, |
||
| 4122 | { |
||
| 4123 | "cell_type": "code", |
||
| 4124 | "execution_count": 241, |
||
| 4125 | "metadata": {}, |
||
| 4126 | "outputs": [ |
||
| 4127 | { |
||
| 4128 | "data": { |
||
| 4129 | "text/plain": [ |
||
| 4130 | "(tensor([[[-0.0504, 0.0757, -0.0366, ..., -0.0275, -0.0294, 0.1300]],\n", |
||
| 4131 | " \n", |
||
| 4132 | " [[-0.0500, 0.0762, -0.0352, ..., -0.0253, -0.0289, 0.1311]],\n", |
||
| 4133 | " \n", |
||
| 4134 | " [[-0.0497, 0.0772, -0.0321, ..., -0.0246, -0.0288, 0.1301]],\n", |
||
| 4135 | " \n", |
||
| 4136 | " [[-0.0491, 0.0794, -0.0337, ..., -0.0260, -0.0278, 0.1281]],\n", |
||
| 4137 | " \n", |
||
| 4138 | " [[-0.0483, 0.0781, -0.0343, ..., -0.0246, -0.0301, 0.1321]],\n", |
||
| 4139 | " \n", |
||
| 4140 | " [[-0.0499, 0.0768, -0.0305, ..., -0.0257, -0.0280, 0.1321]]],\n", |
||
| 4141 | " grad_fn=<AddBackward0>),\n", |
||
| 4142 | " tensor([[[0.0019, 0.0019, 0.0019, ..., 0.0020, 0.0021, 0.0021],\n", |
||
| 4143 | " [0.0017, 0.0020, 0.0020, ..., 0.0019, 0.0019, 0.0018],\n", |
||
| 4144 | " [0.0019, 0.0018, 0.0019, ..., 0.0019, 0.0019, 0.0021],\n", |
||
| 4145 | " [0.0020, 0.0020, 0.0019, ..., 0.0020, 0.0021, 0.0019],\n", |
||
| 4146 | " [0.0017, 0.0023, 0.0021, ..., 0.0019, 0.0020, 0.0020],\n", |
||
| 4147 | " [0.0021, 0.0021, 0.0020, ..., 0.0021, 0.0021, 0.0020]]],\n", |
||
| 4148 | " grad_fn=<DivBackward0>))" |
||
| 4149 | ] |
||
| 4150 | }, |
||
| 4151 | "execution_count": 241, |
||
| 4152 | "metadata": {}, |
||
| 4153 | "output_type": "execute_result" |
||
| 4154 | } |
||
| 4155 | ], |
||
| 4156 | "source": [ |
||
| 4157 | "self.coattn(x_omic, x_path, x_path)" |
||
| 4158 | ] |
||
| 4159 | }, |
||
| 4160 | { |
||
| 4161 | "cell_type": "code", |
||
| 4162 | "execution_count": null, |
||
| 4163 | "metadata": {}, |
||
| 4164 | "outputs": [], |
||
| 4165 | "source": [ |
||
| 4166 | "h" |
||
| 4167 | ] |
||
| 4168 | }, |
||
| 4169 | { |
||
| 4170 | "cell_type": "code", |
||
| 4171 | "execution_count": 208, |
||
| 4172 | "metadata": {}, |
||
| 4173 | "outputs": [], |
||
| 4174 | "source": [ |
||
| 4175 | "sig_feats = [torch.randn(size) for size in sig_sizes]\n", |
||
| 4176 | "x_omic = torch.stack([self.sig_networks[idx].forward(sig_feat) for idx, sig_feat in enumerate(sig_feats)])\n" |
||
| 4177 | ] |
||
| 4178 | }, |
||
| 4179 | { |
||
| 4180 | "cell_type": "code", |
||
| 4181 | "execution_count": 204, |
||
| 4182 | "metadata": {}, |
||
| 4183 | "outputs": [], |
||
| 4184 | "source": [] |
||
| 4185 | }, |
||
| 4186 | { |
||
| 4187 | "cell_type": "code", |
||
| 4188 | "execution_count": 206, |
||
| 4189 | "metadata": {}, |
||
| 4190 | "outputs": [ |
||
| 4191 | { |
||
| 4192 | "data": { |
||
| 4193 | "text/plain": [ |
||
| 4194 | "torch.Size([6, 256])" |
||
| 4195 | ] |
||
| 4196 | }, |
||
| 4197 | "execution_count": 206, |
||
| 4198 | "metadata": {}, |
||
| 4199 | "output_type": "execute_result" |
||
| 4200 | } |
||
| 4201 | ], |
||
| 4202 | "source": [ |
||
| 4203 | "x_omic.shape" |
||
| 4204 | ] |
||
| 4205 | }, |
||
| 4206 | { |
||
| 4207 | "cell_type": "code", |
||
| 4208 | "execution_count": 166, |
||
| 4209 | "metadata": {}, |
||
| 4210 | "outputs": [ |
||
| 4211 | { |
||
| 4212 | "ename": "NameError", |
||
| 4213 | "evalue": "name 'sig1' is not defined", |
||
| 4214 | "output_type": "error", |
||
| 4215 | "traceback": [ |
||
| 4216 | "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", |
||
| 4217 | "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", |
||
| 4218 | "\u001b[0;32m<ipython-input-166-aea4cb4c555c>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0msig1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msig2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msig3\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msig4\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msig5\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msig6\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m", |
||
| 4219 | "\u001b[0;31mNameError\u001b[0m: name 'sig1' is not defined" |
||
| 4220 | ] |
||
| 4221 | } |
||
| 4222 | ], |
||
| 4223 | "source": [ |
||
| 4224 | "sig1, sig2, sig3, sig4, sig5, sig6 = torch.randn()" |
||
| 4225 | ] |
||
| 4226 | }, |
||
| 4227 | { |
||
| 4228 | "cell_type": "code", |
||
| 4229 | "execution_count": 158, |
||
| 4230 | "metadata": {}, |
||
| 4231 | "outputs": [], |
||
| 4232 | "source": [ |
||
| 4233 | "src = torch.rand(6, 1, 256)\n", |
||
| 4234 | "out = transformer(src)\n", |
||
| 4235 | "out = out.squeeze(1).T.unsqueeze(0)" |
||
| 4236 | ] |
||
| 4237 | }, |
||
| 4238 | { |
||
| 4239 | "cell_type": "code", |
||
| 4240 | "execution_count": 163, |
||
| 4241 | "metadata": {}, |
||
| 4242 | "outputs": [], |
||
| 4243 | "source": [ |
||
| 4244 | "conv = nn.Conv1d(in_channels=256, out_channels=256, kernel_size=4, stride=4)" |
||
| 4245 | ] |
||
| 4246 | }, |
||
| 4247 | { |
||
| 4248 | "cell_type": "code", |
||
| 4249 | "execution_count": 164, |
||
| 4250 | "metadata": {}, |
||
| 4251 | "outputs": [ |
||
| 4252 | { |
||
| 4253 | "data": { |
||
| 4254 | "text/plain": [ |
||
| 4255 | "torch.Size([1, 256, 6])" |
||
| 4256 | ] |
||
| 4257 | }, |
||
| 4258 | "execution_count": 164, |
||
| 4259 | "metadata": {}, |
||
| 4260 | "output_type": "execute_result" |
||
| 4261 | } |
||
| 4262 | ], |
||
| 4263 | "source": [ |
||
| 4264 | "out.shape" |
||
| 4265 | ] |
||
| 4266 | }, |
||
| 4267 | { |
||
| 4268 | "cell_type": "code", |
||
| 4269 | "execution_count": 165, |
||
| 4270 | "metadata": {}, |
||
| 4271 | "outputs": [ |
||
| 4272 | { |
||
| 4273 | "data": { |
||
| 4274 | "text/plain": [ |
||
| 4275 | "torch.Size([1, 256, 1])" |
||
| 4276 | ] |
||
| 4277 | }, |
||
| 4278 | "execution_count": 165, |
||
| 4279 | "metadata": {}, |
||
| 4280 | "output_type": "execute_result" |
||
| 4281 | } |
||
| 4282 | ], |
||
| 4283 | "source": [ |
||
| 4284 | "conv(out).shape" |
||
| 4285 | ] |
||
| 4286 | }, |
||
| 4287 | { |
||
| 4288 | "cell_type": "code", |
||
| 4289 | "execution_count": 112, |
||
| 4290 | "metadata": {}, |
||
| 4291 | "outputs": [ |
||
| 4292 | { |
||
| 4293 | "data": { |
||
| 4294 | "text/plain": [ |
||
| 4295 | "torch.Size([1536])" |
||
| 4296 | ] |
||
| 4297 | }, |
||
| 4298 | "execution_count": 112, |
||
| 4299 | "metadata": {}, |
||
| 4300 | "output_type": "execute_result" |
||
| 4301 | } |
||
| 4302 | ], |
||
| 4303 | "source": [ |
||
| 4304 | "x.reshape(-1).shape" |
||
| 4305 | ] |
||
| 4306 | }, |
||
| 4307 | { |
||
| 4308 | "cell_type": "code", |
||
| 4309 | "execution_count": 106, |
||
| 4310 | "metadata": {}, |
||
| 4311 | "outputs": [ |
||
| 4312 | { |
||
| 4313 | "data": { |
||
| 4314 | "text/plain": [ |
||
| 4315 | "3072" |
||
| 4316 | ] |
||
| 4317 | }, |
||
| 4318 | "execution_count": 106, |
||
| 4319 | "metadata": {}, |
||
| 4320 | "output_type": "execute_result" |
||
| 4321 | } |
||
| 4322 | ], |
||
| 4323 | "source": [ |
||
| 4324 | "256 * 12" |
||
| 4325 | ] |
||
| 4326 | }, |
||
| 4327 | { |
||
| 4328 | "cell_type": "code", |
||
| 4329 | "execution_count": 88, |
||
| 4330 | "metadata": {}, |
||
| 4331 | "outputs": [], |
||
| 4332 | "source": [ |
||
| 4333 | "net = Attn_Net_Gated()\n", |
||
| 4334 | "wsi_feats = torch.randn(500, 1, 256)\n", |
||
| 4335 | "sig_feats = torch.randn(6, 1, 256)" |
||
| 4336 | ] |
||
| 4337 | }, |
||
| 4338 | { |
||
| 4339 | "cell_type": "code", |
||
| 4340 | "execution_count": 89, |
||
| 4341 | "metadata": {}, |
||
| 4342 | "outputs": [], |
||
| 4343 | "source": [ |
||
| 4344 | "multihead_attn = nn.MultiheadAttention(embed_dim=256, num_heads=8)" |
||
| 4345 | ] |
||
| 4346 | }, |
||
| 4347 | { |
||
| 4348 | "cell_type": "code", |
||
| 4349 | "execution_count": 90, |
||
| 4350 | "metadata": {}, |
||
| 4351 | "outputs": [], |
||
| 4352 | "source": [ |
||
| 4353 | "out, coattn_weights = multihead_attn(sig_feats, wsi_feats, wsi_feats)" |
||
| 4354 | ] |
||
| 4355 | }, |
||
| 4356 | { |
||
| 4357 | "cell_type": "code", |
||
| 4358 | "execution_count": 96, |
||
| 4359 | "metadata": {}, |
||
| 4360 | "outputs": [], |
||
| 4361 | "source": [ |
||
| 4362 | "cotton = DenseCoAttn(dim1=256, dim2=256, num_attn=8, num_none=3, dropout=0.3b)" |
||
| 4363 | ] |
||
| 4364 | }, |
||
| 4365 | { |
||
| 4366 | "cell_type": "code", |
||
| 4367 | "execution_count": 100, |
||
| 4368 | "metadata": {}, |
||
| 4369 | "outputs": [], |
||
| 4370 | "source": [ |
||
| 4371 | "from math import sqrt\n", |
||
| 4372 | "wsi_feats = torch.randn(1, 500, 256)\n", |
||
| 4373 | "sig_feats = torch.randn(1, 6, 256)\n", |
||
| 4374 | "_ = cotton(wsi_feats, sig_feats)" |
||
| 4375 | ] |
||
| 4376 | }, |
||
| 4377 | { |
||
| 4378 | "cell_type": "code", |
||
| 4379 | "execution_count": 103, |
||
| 4380 | "metadata": {}, |
||
| 4381 | "outputs": [ |
||
| 4382 | { |
||
| 4383 | "data": { |
||
| 4384 | "text/plain": [ |
||
| 4385 | "torch.Size([1, 6, 256])" |
||
| 4386 | ] |
||
| 4387 | }, |
||
| 4388 | "execution_count": 103, |
||
| 4389 | "metadata": {}, |
||
| 4390 | "output_type": "execute_result" |
||
| 4391 | } |
||
| 4392 | ], |
||
| 4393 | "source": [ |
||
| 4394 | "_[0].shape" |
||
| 4395 | ] |
||
| 4396 | }, |
||
| 4397 | { |
||
| 4398 | "cell_type": "code", |
||
| 4399 | "execution_count": 104, |
||
| 4400 | "metadata": {}, |
||
| 4401 | "outputs": [ |
||
| 4402 | { |
||
| 4403 | "data": { |
||
| 4404 | "text/plain": [ |
||
| 4405 | "torch.Size([1, 500, 256])" |
||
| 4406 | ] |
||
| 4407 | }, |
||
| 4408 | "execution_count": 104, |
||
| 4409 | "metadata": {}, |
||
| 4410 | "output_type": "execute_result" |
||
| 4411 | } |
||
| 4412 | ], |
||
| 4413 | "source": [ |
||
| 4414 | "_[1].shape" |
||
| 4415 | ] |
||
| 4416 | }, |
||
| 4417 | { |
||
| 4418 | "cell_type": "code", |
||
| 4419 | "execution_count": 94, |
||
| 4420 | "metadata": {}, |
||
| 4421 | "outputs": [], |
||
| 4422 | "source": [ |
||
| 4423 | "\n", |
||
| 4424 | "import torch\n", |
||
| 4425 | "import torch.nn as nn\n", |
||
| 4426 | "import torch.nn.functional as F\n", |
||
| 4427 | "\n", |
||
| 4428 | "\n", |
||
| 4429 | "def qkv_attention(query, key, value, mask=None, dropout=None):\n", |
||
| 4430 | "\td_k = query.size(-1)\n", |
||
| 4431 | "\tscores = torch.matmul(query, key.transpose(-2,-1)) / sqrt(d_k)\n", |
||
| 4432 | "\tif mask is not None:\n", |
||
| 4433 | "\t\tscores.data.masked_fill_(mask.eq(0), -65504.0)\n", |
||
| 4434 | "\t\n", |
||
| 4435 | "\tp_attn = F.softmax(scores, dim=-1)\n", |
||
| 4436 | "\tif dropout is not None:\n", |
||
| 4437 | "\t\tp_attn = dropout(p_attn)\n", |
||
| 4438 | "\n", |
||
| 4439 | "\treturn torch.matmul(p_attn, value), p_attn\n", |
||
| 4440 | "\n", |
||
| 4441 | "\n", |
||
| 4442 | "class DenseCoAttn(nn.Module):\n", |
||
| 4443 | "\n", |
||
| 4444 | "\tdef __init__(self, dim1, dim2, num_attn, num_none, dropout, is_multi_head=False):\n", |
||
| 4445 | "\t\tsuper(DenseCoAttn, self).__init__()\n", |
||
| 4446 | "\t\tdim = min(dim1, dim2)\n", |
||
| 4447 | "\t\tself.linears = nn.ModuleList([nn.Linear(dim1, dim, bias=False),\n", |
||
| 4448 | "\t\t\t\t\t\t\t\t\t nn.Linear(dim2, dim, bias=False)])\n", |
||
| 4449 | "\t\tself.nones = nn.ParameterList([nn.Parameter(nn.init.xavier_uniform_(torch.empty(num_none, dim1))),\n", |
||
| 4450 | "\t\t\t\t\t\t\t\t\t nn.Parameter(nn.init.xavier_uniform_(torch.empty(num_none, dim2)))])\n", |
||
| 4451 | "\t\tself.d_k = dim // num_attn\n", |
||
| 4452 | "\t\tself.h = num_attn\n", |
||
| 4453 | "\t\tself.num_none = num_none\n", |
||
| 4454 | "\t\tself.is_multi_head = is_multi_head\n", |
||
| 4455 | "\t\tself.attn = None\n", |
||
| 4456 | "\t\tself.dropouts = nn.ModuleList([nn.Dropout(p=dropout) for _ in range(2)])\n", |
||
| 4457 | "\n", |
||
| 4458 | "\tdef forward(self, value1, value2, mask1=None, mask2=None):\n", |
||
| 4459 | "\t\tbatch = value1.size(0)\n", |
||
| 4460 | "\t\tdim1, dim2 = value1.size(-1), value2.size(-1)\n", |
||
| 4461 | "\t\tvalue1 = torch.cat([self.nones[0].unsqueeze(0).expand(batch, self.num_none, dim1), value1], dim=1)\n", |
||
| 4462 | "\t\tvalue2 = torch.cat([self.nones[1].unsqueeze(0).expand(batch, self.num_none, dim2), value2], dim=1)\n", |
||
| 4463 | "\t\tnone_mask = value1.new_ones((batch, self.num_none))\n", |
||
| 4464 | "\n", |
||
| 4465 | "\t\tif mask1 is not None:\n", |
||
| 4466 | "\t\t\tmask1 = torch.cat([none_mask, mask1], dim=1)\n", |
||
| 4467 | "\t\t\tmask1 = mask1.unsqueeze(1).unsqueeze(2)\n", |
||
| 4468 | "\t\tif mask2 is not None:\n", |
||
| 4469 | "\t\t\tmask2 = torch.cat([none_mask, mask2], dim=1)\n", |
||
| 4470 | "\t\t\tmask2 = mask2.unsqueeze(1).unsqueeze(2)\n", |
||
| 4471 | "\n", |
||
| 4472 | "\t\tquery1, query2 = [l(x).view(batch, -1, self.h, self.d_k).transpose(1, 2) \n", |
||
| 4473 | "\t\t\tfor l, x in zip(self.linears, (value1, value2))]\n", |
||
| 4474 | "\n", |
||
| 4475 | "\t\tif self.is_multi_head:\n", |
||
| 4476 | "\t\t\tweighted1, attn1 = qkv_attention(query2, query1, query1, mask=mask1, dropout=self.dropouts[0])\n", |
||
| 4477 | "\t\t\tweighted1 = weighted1.transpose(1, 2).contiguous()[:, self.num_none:, :]\n", |
||
| 4478 | "\t\t\tweighted2, attn2 = qkv_attention(query1, query2, query2, mask=mask2, dropout=self.dropouts[1])\n", |
||
| 4479 | "\t\t\tweighted2 = weighted2.transpose(1, 2).contiguous()[:, self.num_none:, :]\n", |
||
| 4480 | "\t\telse:\n", |
||
| 4481 | "\t\t\tweighted1, attn1 = qkv_attention(query2, query1, value1.unsqueeze(1), mask=mask1, \n", |
||
| 4482 | "\t\t\t\tdropout=self.dropouts[0])\n", |
||
| 4483 | "\t\t\tweighted1 = weighted1.mean(dim=1)[:, self.num_none:, :]\n", |
||
| 4484 | "\t\t\tweighted2, attn2 = qkv_attention(query1, query2, value2.unsqueeze(1), mask=mask2, \n", |
||
| 4485 | "\t\t\t\tdropout=self.dropouts[1])\n", |
||
| 4486 | "\t\t\tweighted2 = weighted2.mean(dim=1)[:, self.num_none:, :]\n", |
||
| 4487 | "\t\tself.attn = [attn1[:,:,self.num_none:,self.num_none:], attn2[:,:,self.num_none:,self.num_none:]]\n", |
||
| 4488 | "\n", |
||
| 4489 | "\t\treturn weighted1, weighted2\n" |
||
| 4490 | ] |
||
| 4491 | }, |
||
| 4492 | { |
||
| 4493 | "cell_type": "code", |
||
| 4494 | "execution_count": null, |
||
| 4495 | "metadata": {}, |
||
| 4496 | "outputs": [], |
||
| 4497 | "source": [] |
||
| 4498 | }, |
||
| 4499 | { |
||
| 4500 | "cell_type": "code", |
||
| 4501 | "execution_count": null, |
||
| 4502 | "metadata": {}, |
||
| 4503 | "outputs": [], |
||
| 4504 | "source": [] |
||
| 4505 | }, |
||
| 4506 | { |
||
| 4507 | "cell_type": "code", |
||
| 4508 | "execution_count": null, |
||
| 4509 | "metadata": {}, |
||
| 4510 | "outputs": [], |
||
| 4511 | "source": [] |
||
| 4512 | }, |
||
| 4513 | { |
||
| 4514 | "cell_type": "code", |
||
| 4515 | "execution_count": 417, |
||
| 4516 | "metadata": {}, |
||
| 4517 | "outputs": [], |
||
| 4518 | "source": [ |
||
| 4519 | "from torch.nn.functional import *\n", |
||
| 4520 | "\n", |
||
| 4521 | "def multi_head_attention_forward(\n", |
||
| 4522 | " query: Tensor,\n", |
||
| 4523 | " key: Tensor,\n", |
||
| 4524 | " value: Tensor,\n", |
||
| 4525 | " embed_dim_to_check: int,\n", |
||
| 4526 | " num_heads: int,\n", |
||
| 4527 | " in_proj_weight: Tensor,\n", |
||
| 4528 | " in_proj_bias: Tensor,\n", |
||
| 4529 | " bias_k: Optional[Tensor],\n", |
||
| 4530 | " bias_v: Optional[Tensor],\n", |
||
| 4531 | " add_zero_attn: bool,\n", |
||
| 4532 | " dropout_p: float,\n", |
||
| 4533 | " out_proj_weight: Tensor,\n", |
||
| 4534 | " out_proj_bias: Tensor,\n", |
||
| 4535 | " training: bool = True,\n", |
||
| 4536 | " key_padding_mask: Optional[Tensor] = None,\n", |
||
| 4537 | " need_weights: bool = True,\n", |
||
| 4538 | " need_raw: bool = True,\n", |
||
| 4539 | " attn_mask: Optional[Tensor] = None,\n", |
||
| 4540 | " use_separate_proj_weight: bool = False,\n", |
||
| 4541 | " q_proj_weight: Optional[Tensor] = None,\n", |
||
| 4542 | " k_proj_weight: Optional[Tensor] = None,\n", |
||
| 4543 | " v_proj_weight: Optional[Tensor] = None,\n", |
||
| 4544 | " static_k: Optional[Tensor] = None,\n", |
||
| 4545 | " static_v: Optional[Tensor] = None,\n", |
||
| 4546 | ") -> Tuple[Tensor, Optional[Tensor]]:\n", |
||
| 4547 | " r\"\"\"\n", |
||
| 4548 | " Args:\n", |
||
| 4549 | " query, key, value: map a query and a set of key-value pairs to an output.\n", |
||
| 4550 | " See \"Attention Is All You Need\" for more details.\n", |
||
| 4551 | " embed_dim_to_check: total dimension of the model.\n", |
||
| 4552 | " num_heads: parallel attention heads.\n", |
||
| 4553 | " in_proj_weight, in_proj_bias: input projection weight and bias.\n", |
||
| 4554 | " bias_k, bias_v: bias of the key and value sequences to be added at dim=0.\n", |
||
| 4555 | " add_zero_attn: add a new batch of zeros to the key and\n", |
||
| 4556 | " value sequences at dim=1.\n", |
||
| 4557 | " dropout_p: probability of an element to be zeroed.\n", |
||
| 4558 | " out_proj_weight, out_proj_bias: the output projection weight and bias.\n", |
||
| 4559 | " training: apply dropout if is ``True``.\n", |
||
| 4560 | " key_padding_mask: if provided, specified padding elements in the key will\n", |
||
| 4561 | " be ignored by the attention. This is an binary mask. When the value is True,\n", |
||
| 4562 | " the corresponding value on the attention layer will be filled with -inf.\n", |
||
| 4563 | " need_weights: output attn_output_weights.\n", |
||
| 4564 | " attn_mask: 2D or 3D mask that prevents attention to certain positions. A 2D mask will be broadcasted for all\n", |
||
| 4565 | " the batches while a 3D mask allows to specify a different mask for the entries of each batch.\n", |
||
| 4566 | " use_separate_proj_weight: the function accept the proj. weights for query, key,\n", |
||
| 4567 | " and value in different forms. If false, in_proj_weight will be used, which is\n", |
||
| 4568 | " a combination of q_proj_weight, k_proj_weight, v_proj_weight.\n", |
||
| 4569 | " q_proj_weight, k_proj_weight, v_proj_weight, in_proj_bias: input projection weight and bias.\n", |
||
| 4570 | " static_k, static_v: static key and value used for attention operators.\n", |
||
| 4571 | " Shape:\n", |
||
| 4572 | " Inputs:\n", |
||
| 4573 | " - query: :math:`(L, N, E)` where L is the target sequence length, N is the batch size, E is\n", |
||
| 4574 | " the embedding dimension.\n", |
||
| 4575 | " - key: :math:`(S, N, E)`, where S is the source sequence length, N is the batch size, E is\n", |
||
| 4576 | " the embedding dimension.\n", |
||
| 4577 | " - value: :math:`(S, N, E)` where S is the source sequence length, N is the batch size, E is\n", |
||
| 4578 | " the embedding dimension.\n", |
||
| 4579 | " - key_padding_mask: :math:`(N, S)` where N is the batch size, S is the source sequence length.\n", |
||
| 4580 | " If a ByteTensor is provided, the non-zero positions will be ignored while the zero positions\n", |
||
| 4581 | " will be unchanged. If a BoolTensor is provided, the positions with the\n", |
||
| 4582 | " value of ``True`` will be ignored while the position with the value of ``False`` will be unchanged.\n", |
||
| 4583 | " - attn_mask: 2D mask :math:`(L, S)` where L is the target sequence length, S is the source sequence length.\n", |
||
| 4584 | " 3D mask :math:`(N*num_heads, L, S)` where N is the batch size, L is the target sequence length,\n", |
||
| 4585 | " S is the source sequence length. attn_mask ensures that position i is allowed to attend the unmasked\n", |
||
| 4586 | " positions. If a ByteTensor is provided, the non-zero positions are not allowed to attend\n", |
||
| 4587 | " while the zero positions will be unchanged. If a BoolTensor is provided, positions with ``True``\n", |
||
| 4588 | " are not allowed to attend while ``False`` values will be unchanged. If a FloatTensor\n", |
||
| 4589 | " is provided, it will be added to the attention weight.\n", |
||
| 4590 | " - static_k: :math:`(N*num_heads, S, E/num_heads)`, where S is the source sequence length,\n", |
||
| 4591 | " N is the batch size, E is the embedding dimension. E/num_heads is the head dimension.\n", |
||
| 4592 | " - static_v: :math:`(N*num_heads, S, E/num_heads)`, where S is the source sequence length,\n", |
||
| 4593 | " N is the batch size, E is the embedding dimension. E/num_heads is the head dimension.\n", |
||
| 4594 | " Outputs:\n", |
||
| 4595 | " - attn_output: :math:`(L, N, E)` where L is the target sequence length, N is the batch size,\n", |
||
| 4596 | " E is the embedding dimension.\n", |
||
| 4597 | " - attn_output_weights: :math:`(N, L, S)` where N is the batch size,\n", |
||
| 4598 | " L is the target sequence length, S is the source sequence length.\n", |
||
| 4599 | " \"\"\"\n", |
||
| 4600 | " tens_ops = (query, key, value, in_proj_weight, in_proj_bias, bias_k, bias_v, out_proj_weight, out_proj_bias)\n", |
||
| 4601 | " if has_torch_function(tens_ops):\n", |
||
| 4602 | " return handle_torch_function(\n", |
||
| 4603 | " multi_head_attention_forward,\n", |
||
| 4604 | " tens_ops,\n", |
||
| 4605 | " query,\n", |
||
| 4606 | " key,\n", |
||
| 4607 | " value,\n", |
||
| 4608 | " embed_dim_to_check,\n", |
||
| 4609 | " num_heads,\n", |
||
| 4610 | " in_proj_weight,\n", |
||
| 4611 | " in_proj_bias,\n", |
||
| 4612 | " bias_k,\n", |
||
| 4613 | " bias_v,\n", |
||
| 4614 | " add_zero_attn,\n", |
||
| 4615 | " dropout_p,\n", |
||
| 4616 | " out_proj_weight,\n", |
||
| 4617 | " out_proj_bias,\n", |
||
| 4618 | " training=training,\n", |
||
| 4619 | " key_padding_mask=key_padding_mask,\n", |
||
| 4620 | " need_weights=need_weights,\n", |
||
| 4621 | " need_raw=need_raw,\n", |
||
| 4622 | " attn_mask=attn_mask,\n", |
||
| 4623 | " use_separate_proj_weight=use_separate_proj_weight,\n", |
||
| 4624 | " q_proj_weight=q_proj_weight,\n", |
||
| 4625 | " k_proj_weight=k_proj_weight,\n", |
||
| 4626 | " v_proj_weight=v_proj_weight,\n", |
||
| 4627 | " static_k=static_k,\n", |
||
| 4628 | " static_v=static_v,\n", |
||
| 4629 | " )\n", |
||
| 4630 | " tgt_len, bsz, embed_dim = query.size()\n", |
||
| 4631 | " assert embed_dim == embed_dim_to_check\n", |
||
| 4632 | " # allow MHA to have different sizes for the feature dimension\n", |
||
| 4633 | " assert key.size(0) == value.size(0) and key.size(1) == value.size(1)\n", |
||
| 4634 | "\n", |
||
| 4635 | " head_dim = embed_dim // num_heads\n", |
||
| 4636 | " assert head_dim * num_heads == embed_dim, \"embed_dim must be divisible by num_heads\"\n", |
||
| 4637 | " scaling = float(head_dim) ** -0.5\n", |
||
| 4638 | "\n", |
||
| 4639 | " if not use_separate_proj_weight:\n", |
||
| 4640 | " if (query is key or torch.equal(query, key)) and (key is value or torch.equal(key, value)):\n", |
||
| 4641 | " # self-attention\n", |
||
| 4642 | " q, k, v = linear(query, in_proj_weight, in_proj_bias).chunk(3, dim=-1)\n", |
||
| 4643 | "\n", |
||
| 4644 | " elif key is value or torch.equal(key, value):\n", |
||
| 4645 | " # encoder-decoder attention\n", |
||
| 4646 | " # This is inline in_proj function with in_proj_weight and in_proj_bias\n", |
||
| 4647 | " _b = in_proj_bias\n", |
||
| 4648 | " _start = 0\n", |
||
| 4649 | " _end = embed_dim\n", |
||
| 4650 | " _w = in_proj_weight[_start:_end, :]\n", |
||
| 4651 | " if _b is not None:\n", |
||
| 4652 | " _b = _b[_start:_end]\n", |
||
| 4653 | " q = linear(query, _w, _b)\n", |
||
| 4654 | "\n", |
||
| 4655 | " if key is None:\n", |
||
| 4656 | " assert value is None\n", |
||
| 4657 | " k = None\n", |
||
| 4658 | " v = None\n", |
||
| 4659 | " else:\n", |
||
| 4660 | "\n", |
||
| 4661 | " # This is inline in_proj function with in_proj_weight and in_proj_bias\n", |
||
| 4662 | " _b = in_proj_bias\n", |
||
| 4663 | " _start = embed_dim\n", |
||
| 4664 | " _end = None\n", |
||
| 4665 | " _w = in_proj_weight[_start:, :]\n", |
||
| 4666 | " if _b is not None:\n", |
||
| 4667 | " _b = _b[_start:]\n", |
||
| 4668 | " k, v = linear(key, _w, _b).chunk(2, dim=-1)\n", |
||
| 4669 | "\n", |
||
| 4670 | " else:\n", |
||
| 4671 | " # This is inline in_proj function with in_proj_weight and in_proj_bias\n", |
||
| 4672 | " _b = in_proj_bias\n", |
||
| 4673 | " _start = 0\n", |
||
| 4674 | " _end = embed_dim\n", |
||
| 4675 | " _w = in_proj_weight[_start:_end, :]\n", |
||
| 4676 | " if _b is not None:\n", |
||
| 4677 | " _b = _b[_start:_end]\n", |
||
| 4678 | " q = linear(query, _w, _b)\n", |
||
| 4679 | "\n", |
||
| 4680 | " # This is inline in_proj function with in_proj_weight and in_proj_bias\n", |
||
| 4681 | " _b = in_proj_bias\n", |
||
| 4682 | " _start = embed_dim\n", |
||
| 4683 | " _end = embed_dim * 2\n", |
||
| 4684 | " _w = in_proj_weight[_start:_end, :]\n", |
||
| 4685 | " if _b is not None:\n", |
||
| 4686 | " _b = _b[_start:_end]\n", |
||
| 4687 | " k = linear(key, _w, _b)\n", |
||
| 4688 | "\n", |
||
| 4689 | " # This is inline in_proj function with in_proj_weight and in_proj_bias\n", |
||
| 4690 | " _b = in_proj_bias\n", |
||
| 4691 | " _start = embed_dim * 2\n", |
||
| 4692 | " _end = None\n", |
||
| 4693 | " _w = in_proj_weight[_start:, :]\n", |
||
| 4694 | " if _b is not None:\n", |
||
| 4695 | " _b = _b[_start:]\n", |
||
| 4696 | " v = linear(value, _w, _b)\n", |
||
| 4697 | " else:\n", |
||
| 4698 | " q_proj_weight_non_opt = torch.jit._unwrap_optional(q_proj_weight)\n", |
||
| 4699 | " len1, len2 = q_proj_weight_non_opt.size()\n", |
||
| 4700 | " assert len1 == embed_dim and len2 == query.size(-1)\n", |
||
| 4701 | "\n", |
||
| 4702 | " k_proj_weight_non_opt = torch.jit._unwrap_optional(k_proj_weight)\n", |
||
| 4703 | " len1, len2 = k_proj_weight_non_opt.size()\n", |
||
| 4704 | " assert len1 == embed_dim and len2 == key.size(-1)\n", |
||
| 4705 | "\n", |
||
| 4706 | " v_proj_weight_non_opt = torch.jit._unwrap_optional(v_proj_weight)\n", |
||
| 4707 | " len1, len2 = v_proj_weight_non_opt.size()\n", |
||
| 4708 | " assert len1 == embed_dim and len2 == value.size(-1)\n", |
||
| 4709 | "\n", |
||
| 4710 | " if in_proj_bias is not None:\n", |
||
| 4711 | " q = linear(query, q_proj_weight_non_opt, in_proj_bias[0:embed_dim])\n", |
||
| 4712 | " k = linear(key, k_proj_weight_non_opt, in_proj_bias[embed_dim : (embed_dim * 2)])\n", |
||
| 4713 | " v = linear(value, v_proj_weight_non_opt, in_proj_bias[(embed_dim * 2) :])\n", |
||
| 4714 | " else:\n", |
||
| 4715 | " q = linear(query, q_proj_weight_non_opt, in_proj_bias)\n", |
||
| 4716 | " k = linear(key, k_proj_weight_non_opt, in_proj_bias)\n", |
||
| 4717 | " v = linear(value, v_proj_weight_non_opt, in_proj_bias)\n", |
||
| 4718 | " q = q * scaling\n", |
||
| 4719 | "\n", |
||
| 4720 | " if attn_mask is not None:\n", |
||
| 4721 | " assert (\n", |
||
| 4722 | " attn_mask.dtype == torch.float32\n", |
||
| 4723 | " or attn_mask.dtype == torch.float64\n", |
||
| 4724 | " or attn_mask.dtype == torch.float16\n", |
||
| 4725 | " or attn_mask.dtype == torch.uint8\n", |
||
| 4726 | " or attn_mask.dtype == torch.bool\n", |
||
| 4727 | " ), \"Only float, byte, and bool types are supported for attn_mask, not {}\".format(attn_mask.dtype)\n", |
||
| 4728 | " if attn_mask.dtype == torch.uint8:\n", |
||
| 4729 | " warnings.warn(\"Byte tensor for attn_mask in nn.MultiheadAttention is deprecated. Use bool tensor instead.\")\n", |
||
| 4730 | " attn_mask = attn_mask.to(torch.bool)\n", |
||
| 4731 | "\n", |
||
| 4732 | " if attn_mask.dim() == 2:\n", |
||
| 4733 | " attn_mask = attn_mask.unsqueeze(0)\n", |
||
| 4734 | " if list(attn_mask.size()) != [1, query.size(0), key.size(0)]:\n", |
||
| 4735 | " raise RuntimeError(\"The size of the 2D attn_mask is not correct.\")\n", |
||
| 4736 | " elif attn_mask.dim() == 3:\n", |
||
| 4737 | " if list(attn_mask.size()) != [bsz * num_heads, query.size(0), key.size(0)]:\n", |
||
| 4738 | " raise RuntimeError(\"The size of the 3D attn_mask is not correct.\")\n", |
||
| 4739 | " else:\n", |
||
| 4740 | " raise RuntimeError(\"attn_mask's dimension {} is not supported\".format(attn_mask.dim()))\n", |
||
| 4741 | " # attn_mask's dim is 3 now.\n", |
||
| 4742 | "\n", |
||
| 4743 | " # convert ByteTensor key_padding_mask to bool\n", |
||
| 4744 | " if key_padding_mask is not None and key_padding_mask.dtype == torch.uint8:\n", |
||
| 4745 | " warnings.warn(\n", |
||
| 4746 | " \"Byte tensor for key_padding_mask in nn.MultiheadAttention is deprecated. Use bool tensor instead.\"\n", |
||
| 4747 | " )\n", |
||
| 4748 | " key_padding_mask = key_padding_mask.to(torch.bool)\n", |
||
| 4749 | "\n", |
||
| 4750 | " if bias_k is not None and bias_v is not None:\n", |
||
| 4751 | " if static_k is None and static_v is None:\n", |
||
| 4752 | " k = torch.cat([k, bias_k.repeat(1, bsz, 1)])\n", |
||
| 4753 | " v = torch.cat([v, bias_v.repeat(1, bsz, 1)])\n", |
||
| 4754 | " if attn_mask is not None:\n", |
||
| 4755 | " attn_mask = pad(attn_mask, (0, 1))\n", |
||
| 4756 | " if key_padding_mask is not None:\n", |
||
| 4757 | " key_padding_mask = pad(key_padding_mask, (0, 1))\n", |
||
| 4758 | " else:\n", |
||
| 4759 | " assert static_k is None, \"bias cannot be added to static key.\"\n", |
||
| 4760 | " assert static_v is None, \"bias cannot be added to static value.\"\n", |
||
| 4761 | " else:\n", |
||
| 4762 | " assert bias_k is None\n", |
||
| 4763 | " assert bias_v is None\n", |
||
| 4764 | "\n", |
||
| 4765 | " q = q.contiguous().view(tgt_len, bsz * num_heads, head_dim).transpose(0, 1)\n", |
||
| 4766 | " if k is not None:\n", |
||
| 4767 | " k = k.contiguous().view(-1, bsz * num_heads, head_dim).transpose(0, 1)\n", |
||
| 4768 | " if v is not None:\n", |
||
| 4769 | " v = v.contiguous().view(-1, bsz * num_heads, head_dim).transpose(0, 1)\n", |
||
| 4770 | "\n", |
||
| 4771 | " if static_k is not None:\n", |
||
| 4772 | " assert static_k.size(0) == bsz * num_heads\n", |
||
| 4773 | " assert static_k.size(2) == head_dim\n", |
||
| 4774 | " k = static_k\n", |
||
| 4775 | "\n", |
||
| 4776 | " if static_v is not None:\n", |
||
| 4777 | " assert static_v.size(0) == bsz * num_heads\n", |
||
| 4778 | " assert static_v.size(2) == head_dim\n", |
||
| 4779 | " v = static_v\n", |
||
| 4780 | "\n", |
||
| 4781 | " src_len = k.size(1)\n", |
||
| 4782 | "\n", |
||
| 4783 | " if key_padding_mask is not None:\n", |
||
| 4784 | " assert key_padding_mask.size(0) == bsz\n", |
||
| 4785 | " assert key_padding_mask.size(1) == src_len\n", |
||
| 4786 | "\n", |
||
| 4787 | " if add_zero_attn:\n", |
||
| 4788 | " src_len += 1\n", |
||
| 4789 | " k = torch.cat([k, torch.zeros((k.size(0), 1) + k.size()[2:], dtype=k.dtype, device=k.device)], dim=1)\n", |
||
| 4790 | " v = torch.cat([v, torch.zeros((v.size(0), 1) + v.size()[2:], dtype=v.dtype, device=v.device)], dim=1)\n", |
||
| 4791 | " if attn_mask is not None:\n", |
||
| 4792 | " attn_mask = pad(attn_mask, (0, 1))\n", |
||
| 4793 | " if key_padding_mask is not None:\n", |
||
| 4794 | " key_padding_mask = pad(key_padding_mask, (0, 1))\n", |
||
| 4795 | "\n", |
||
| 4796 | " attn_output_weights = torch.bmm(q, k.transpose(1, 2))\n", |
||
| 4797 | " assert list(attn_output_weights.size()) == [bsz * num_heads, tgt_len, src_len]\n", |
||
| 4798 | "\n", |
||
| 4799 | " if attn_mask is not None:\n", |
||
| 4800 | " if attn_mask.dtype == torch.bool:\n", |
||
| 4801 | " attn_output_weights.masked_fill_(attn_mask, float(\"-inf\"))\n", |
||
| 4802 | " else:\n", |
||
| 4803 | " attn_output_weights += attn_mask\n", |
||
| 4804 | "\n", |
||
| 4805 | " if key_padding_mask is not None:\n", |
||
| 4806 | " attn_output_weights = attn_output_weights.view(bsz, num_heads, tgt_len, src_len)\n", |
||
| 4807 | " attn_output_weights = attn_output_weights.masked_fill(\n", |
||
| 4808 | " key_padding_mask.unsqueeze(1).unsqueeze(2),\n", |
||
| 4809 | " float(\"-inf\"),\n", |
||
| 4810 | " )\n", |
||
| 4811 | " attn_output_weights = attn_output_weights.view(bsz * num_heads, tgt_len, src_len)\n", |
||
| 4812 | " \n", |
||
| 4813 | " attn_output_weights_raw = attn_output_weights\n", |
||
| 4814 | " attn_output_weights = softmax(attn_output_weights, dim=-1)\n", |
||
| 4815 | " attn_output_weights = dropout(attn_output_weights, p=dropout_p, training=training)\n", |
||
| 4816 | "\n", |
||
| 4817 | " attn_output = torch.bmm(attn_output_weights, v)\n", |
||
| 4818 | " assert list(attn_output.size()) == [bsz * num_heads, tgt_len, head_dim]\n", |
||
| 4819 | " attn_output = attn_output.transpose(0, 1).contiguous().view(tgt_len, bsz, embed_dim)\n", |
||
| 4820 | " attn_output = linear(attn_output, out_proj_weight, out_proj_bias)\n", |
||
| 4821 | " \n", |
||
| 4822 | " if need_weights:\n", |
||
| 4823 | " if need_raw:\n", |
||
| 4824 | " \n", |
||
| 4825 | " attn_output_weights_raw = attn_output_weights_raw.view(bsz, num_heads, tgt_len, src_len)\n", |
||
| 4826 | " return attn_output,attn_output_weights_raw\n", |
||
| 4827 | " \n", |
||
| 4828 | " #attn_output_weights = attn_output_weights.view(bsz, num_heads, tgt_len, src_len)\n", |
||
| 4829 | " #return attn_output, attn_output_weights.sum(dim=1) / num_heads, attn_output_weights_raw, attn_output_weights_raw.sum(dim=1) / num_heads\n", |
||
| 4830 | " else:\n", |
||
| 4831 | " # average attention weights over heads\n", |
||
| 4832 | " attn_output_weights = attn_output_weights.view(bsz, num_heads, tgt_len, src_len)\n", |
||
| 4833 | " return attn_output, attn_output_weights.sum(dim=1) / num_heads\n", |
||
| 4834 | " else:\n", |
||
| 4835 | " return attn_output, None\n" |
||
| 4836 | ] |
||
| 4837 | }, |
||
| 4838 | { |
||
| 4839 | "cell_type": "code", |
||
| 4840 | "execution_count": 418, |
||
| 4841 | "metadata": {}, |
||
| 4842 | "outputs": [], |
||
| 4843 | "source": [ |
||
| 4844 | "import torch\n", |
||
| 4845 | "from torch import Tensor\n", |
||
| 4846 | "from torch.nn.modules.linear import _LinearWithBias\n", |
||
| 4847 | "from torch.nn.init import xavier_uniform_\n", |
||
| 4848 | "from torch.nn.init import constant_\n", |
||
| 4849 | "from torch.nn.init import xavier_normal_\n", |
||
| 4850 | "from torch.nn.parameter import Parameter\n", |
||
| 4851 | "from torch.nn import Module\n", |
||
| 4852 | "\n", |
||
| 4853 | "class MultiheadAttention(Module):\n", |
||
| 4854 | " r\"\"\"Allows the model to jointly attend to information\n", |
||
| 4855 | " from different representation subspaces.\n", |
||
| 4856 | " See reference: Attention Is All You Need\n", |
||
| 4857 | "\n", |
||
| 4858 | " .. math::\n", |
||
| 4859 | " \\text{MultiHead}(Q, K, V) = \\text{Concat}(head_1,\\dots,head_h)W^O\n", |
||
| 4860 | " \\text{where} head_i = \\text{Attention}(QW_i^Q, KW_i^K, VW_i^V)\n", |
||
| 4861 | "\n", |
||
| 4862 | " Args:\n", |
||
| 4863 | " embed_dim: total dimension of the model.\n", |
||
| 4864 | " num_heads: parallel attention heads.\n", |
||
| 4865 | " dropout: a Dropout layer on attn_output_weights. Default: 0.0.\n", |
||
| 4866 | " bias: add bias as module parameter. Default: True.\n", |
||
| 4867 | " add_bias_kv: add bias to the key and value sequences at dim=0.\n", |
||
| 4868 | " add_zero_attn: add a new batch of zeros to the key and\n", |
||
| 4869 | " value sequences at dim=1.\n", |
||
| 4870 | " kdim: total number of features in key. Default: None.\n", |
||
| 4871 | " vdim: total number of features in value. Default: None.\n", |
||
| 4872 | "\n", |
||
| 4873 | " Note: if kdim and vdim are None, they will be set to embed_dim such that\n", |
||
| 4874 | " query, key, and value have the same number of features.\n", |
||
| 4875 | "\n", |
||
| 4876 | " Examples::\n", |
||
| 4877 | "\n", |
||
| 4878 | " >>> multihead_attn = nn.MultiheadAttention(embed_dim, num_heads)\n", |
||
| 4879 | " >>> attn_output, attn_output_weights = multihead_attn(query, key, value)\n", |
||
| 4880 | " \"\"\"\n", |
||
| 4881 | " bias_k: Optional[torch.Tensor]\n", |
||
| 4882 | " bias_v: Optional[torch.Tensor]\n", |
||
| 4883 | "\n", |
||
| 4884 | " def __init__(self, embed_dim, num_heads, dropout=0., bias=True, add_bias_kv=False, add_zero_attn=False, kdim=None, vdim=None):\n", |
||
| 4885 | " super(MultiheadAttention, self).__init__()\n", |
||
| 4886 | " self.embed_dim = embed_dim\n", |
||
| 4887 | " self.kdim = kdim if kdim is not None else embed_dim\n", |
||
| 4888 | " self.vdim = vdim if vdim is not None else embed_dim\n", |
||
| 4889 | " self._qkv_same_embed_dim = self.kdim == embed_dim and self.vdim == embed_dim\n", |
||
| 4890 | "\n", |
||
| 4891 | " self.num_heads = num_heads\n", |
||
| 4892 | " self.dropout = dropout\n", |
||
| 4893 | " self.head_dim = embed_dim // num_heads\n", |
||
| 4894 | " assert self.head_dim * num_heads == self.embed_dim, \"embed_dim must be divisible by num_heads\"\n", |
||
| 4895 | "\n", |
||
| 4896 | " if self._qkv_same_embed_dim is False:\n", |
||
| 4897 | " self.q_proj_weight = Parameter(torch.Tensor(embed_dim, embed_dim))\n", |
||
| 4898 | " self.k_proj_weight = Parameter(torch.Tensor(embed_dim, self.kdim))\n", |
||
| 4899 | " self.v_proj_weight = Parameter(torch.Tensor(embed_dim, self.vdim))\n", |
||
| 4900 | " self.register_parameter('in_proj_weight', None)\n", |
||
| 4901 | " else:\n", |
||
| 4902 | " self.in_proj_weight = Parameter(torch.empty(3 * embed_dim, embed_dim))\n", |
||
| 4903 | " self.register_parameter('q_proj_weight', None)\n", |
||
| 4904 | " self.register_parameter('k_proj_weight', None)\n", |
||
| 4905 | " self.register_parameter('v_proj_weight', None)\n", |
||
| 4906 | "\n", |
||
| 4907 | " if bias:\n", |
||
| 4908 | " self.in_proj_bias = Parameter(torch.empty(3 * embed_dim))\n", |
||
| 4909 | " else:\n", |
||
| 4910 | " self.register_parameter('in_proj_bias', None)\n", |
||
| 4911 | " self.out_proj = _LinearWithBias(embed_dim, embed_dim)\n", |
||
| 4912 | "\n", |
||
| 4913 | " if add_bias_kv:\n", |
||
| 4914 | " self.bias_k = Parameter(torch.empty(1, 1, embed_dim))\n", |
||
| 4915 | " self.bias_v = Parameter(torch.empty(1, 1, embed_dim))\n", |
||
| 4916 | " else:\n", |
||
| 4917 | " self.bias_k = self.bias_v = None\n", |
||
| 4918 | "\n", |
||
| 4919 | " self.add_zero_attn = add_zero_attn\n", |
||
| 4920 | "\n", |
||
| 4921 | " self._reset_parameters()\n", |
||
| 4922 | "\n", |
||
| 4923 | " def _reset_parameters(self):\n", |
||
| 4924 | " if self._qkv_same_embed_dim:\n", |
||
| 4925 | " xavier_uniform_(self.in_proj_weight)\n", |
||
| 4926 | " else:\n", |
||
| 4927 | " xavier_uniform_(self.q_proj_weight)\n", |
||
| 4928 | " xavier_uniform_(self.k_proj_weight)\n", |
||
| 4929 | " xavier_uniform_(self.v_proj_weight)\n", |
||
| 4930 | "\n", |
||
| 4931 | " if self.in_proj_bias is not None:\n", |
||
| 4932 | " constant_(self.in_proj_bias, 0.)\n", |
||
| 4933 | " constant_(self.out_proj.bias, 0.)\n", |
||
| 4934 | " if self.bias_k is not None:\n", |
||
| 4935 | " xavier_normal_(self.bias_k)\n", |
||
| 4936 | " if self.bias_v is not None:\n", |
||
| 4937 | " xavier_normal_(self.bias_v)\n", |
||
| 4938 | "\n", |
||
| 4939 | " def __setstate__(self, state):\n", |
||
| 4940 | " # Support loading old MultiheadAttention checkpoints generated by v1.1.0\n", |
||
| 4941 | " if '_qkv_same_embed_dim' not in state:\n", |
||
| 4942 | " state['_qkv_same_embed_dim'] = True\n", |
||
| 4943 | "\n", |
||
| 4944 | " super(MultiheadAttention, self).__setstate__(state)\n", |
||
| 4945 | "\n", |
||
| 4946 | " def forward(self, query, key, value, key_padding_mask=None,\n", |
||
| 4947 | " need_weights=True, need_raw=True, attn_mask=None):\n", |
||
| 4948 | " # type: (Tensor, Tensor, Tensor, Optional[Tensor], bool, Optional[Tensor]) -> Tuple[Tensor, Optional[Tensor]]\n", |
||
| 4949 | " r\"\"\"\n", |
||
| 4950 | " Args:\n", |
||
| 4951 | " query, key, value: map a query and a set of key-value pairs to an output.\n", |
||
| 4952 | " See \"Attention Is All You Need\" for more details.\n", |
||
| 4953 | " key_padding_mask: if provided, specified padding elements in the key will\n", |
||
| 4954 | " be ignored by the attention. When given a binary mask and a value is True,\n", |
||
| 4955 | " the corresponding value on the attention layer will be ignored. When given\n", |
||
| 4956 | " a byte mask and a value is non-zero, the corresponding value on the attention\n", |
||
| 4957 | " layer will be ignored\n", |
||
| 4958 | " need_weights: output attn_output_weights.\n", |
||
| 4959 | " attn_mask: 2D or 3D mask that prevents attention to certain positions. A 2D mask will be broadcasted for all\n", |
||
| 4960 | " the batches while a 3D mask allows to specify a different mask for the entries of each batch.\n", |
||
| 4961 | "\n", |
||
| 4962 | " Shape:\n", |
||
| 4963 | " - Inputs:\n", |
||
| 4964 | " - query: :math:`(L, N, E)` where L is the target sequence length, N is the batch size, E is\n", |
||
| 4965 | " the embedding dimension.\n", |
||
| 4966 | " - key: :math:`(S, N, E)`, where S is the source sequence length, N is the batch size, E is\n", |
||
| 4967 | " the embedding dimension.\n", |
||
| 4968 | " - value: :math:`(S, N, E)` where S is the source sequence length, N is the batch size, E is\n", |
||
| 4969 | " the embedding dimension.\n", |
||
| 4970 | " - key_padding_mask: :math:`(N, S)` where N is the batch size, S is the source sequence length.\n", |
||
| 4971 | " If a ByteTensor is provided, the non-zero positions will be ignored while the position\n", |
||
| 4972 | " with the zero positions will be unchanged. If a BoolTensor is provided, the positions with the\n", |
||
| 4973 | " value of ``True`` will be ignored while the position with the value of ``False`` will be unchanged.\n", |
||
| 4974 | " - attn_mask: 2D mask :math:`(L, S)` where L is the target sequence length, S is the source sequence length.\n", |
||
| 4975 | " 3D mask :math:`(N*num_heads, L, S)` where N is the batch size, L is the target sequence length,\n", |
||
| 4976 | " S is the source sequence length. attn_mask ensure that position i is allowed to attend the unmasked\n", |
||
| 4977 | " positions. If a ByteTensor is provided, the non-zero positions are not allowed to attend\n", |
||
| 4978 | " while the zero positions will be unchanged. If a BoolTensor is provided, positions with ``True``\n", |
||
| 4979 | " is not allowed to attend while ``False`` values will be unchanged. If a FloatTensor\n", |
||
| 4980 | " is provided, it will be added to the attention weight.\n", |
||
| 4981 | "\n", |
||
| 4982 | " - Outputs:\n", |
||
| 4983 | " - attn_output: :math:`(L, N, E)` where L is the target sequence length, N is the batch size,\n", |
||
| 4984 | " E is the embedding dimension.\n", |
||
| 4985 | " - attn_output_weights: :math:`(N, L, S)` where N is the batch size,\n", |
||
| 4986 | " L is the target sequence length, S is the source sequence length.\n", |
||
| 4987 | " \"\"\"\n", |
||
| 4988 | " if not self._qkv_same_embed_dim:\n", |
||
| 4989 | " return multi_head_attention_forward(\n", |
||
| 4990 | " query, key, value, self.embed_dim, self.num_heads,\n", |
||
| 4991 | " self.in_proj_weight, self.in_proj_bias,\n", |
||
| 4992 | " self.bias_k, self.bias_v, self.add_zero_attn,\n", |
||
| 4993 | " self.dropout, self.out_proj.weight, self.out_proj.bias,\n", |
||
| 4994 | " training=self.training,\n", |
||
| 4995 | " key_padding_mask=key_padding_mask, need_weights=need_weights, need_raw=need_raw,\n", |
||
| 4996 | " attn_mask=attn_mask, use_separate_proj_weight=True,\n", |
||
| 4997 | " q_proj_weight=self.q_proj_weight, k_proj_weight=self.k_proj_weight,\n", |
||
| 4998 | " v_proj_weight=self.v_proj_weight)\n", |
||
| 4999 | " else:\n", |
||
| 5000 | " return multi_head_attention_forward(\n", |
||
| 5001 | " query, key, value, self.embed_dim, self.num_heads,\n", |
||
| 5002 | " self.in_proj_weight, self.in_proj_bias,\n", |
||
| 5003 | " self.bias_k, self.bias_v, self.add_zero_attn,\n", |
||
| 5004 | " self.dropout, self.out_proj.weight, self.out_proj.bias,\n", |
||
| 5005 | " training=self.training,\n", |
||
| 5006 | " key_padding_mask=key_padding_mask, need_weights=need_weights, need_raw=need_raw,\n", |
||
| 5007 | " attn_mask=attn_mask)" |
||
| 5008 | ] |
||
| 5009 | }, |
||
| 5010 | { |
||
| 5011 | "cell_type": "code", |
||
| 5012 | "execution_count": null, |
||
| 5013 | "metadata": {}, |
||
| 5014 | "outputs": [], |
||
| 5015 | "source": [] |
||
| 5016 | }, |
||
| 5017 | { |
||
| 5018 | "cell_type": "code", |
||
| 5019 | "execution_count": null, |
||
| 5020 | "metadata": {}, |
||
| 5021 | "outputs": [], |
||
| 5022 | "source": [] |
||
| 5023 | }, |
||
| 5024 | { |
||
| 5025 | "cell_type": "code", |
||
| 5026 | "execution_count": null, |
||
| 5027 | "metadata": {}, |
||
| 5028 | "outputs": [], |
||
| 5029 | "source": [] |
||
| 5030 | }, |
||
| 5031 | { |
||
| 5032 | "cell_type": "code", |
||
| 5033 | "execution_count": null, |
||
| 5034 | "metadata": {}, |
||
| 5035 | "outputs": [], |
||
| 5036 | "source": [] |
||
| 5037 | }, |
||
| 5038 | { |
||
| 5039 | "cell_type": "code", |
||
| 5040 | "execution_count": null, |
||
| 5041 | "metadata": {}, |
||
| 5042 | "outputs": [], |
||
| 5043 | "source": [] |
||
| 5044 | }, |
||
| 5045 | { |
||
| 5046 | "cell_type": "code", |
||
| 5047 | "execution_count": null, |
||
| 5048 | "metadata": {}, |
||
| 5049 | "outputs": [], |
||
| 5050 | "source": [] |
||
| 5051 | }, |
||
| 5052 | { |
||
| 5053 | "cell_type": "code", |
||
| 5054 | "execution_count": null, |
||
| 5055 | "metadata": {}, |
||
| 5056 | "outputs": [], |
||
| 5057 | "source": [] |
||
| 5058 | }, |
||
| 5059 | { |
||
| 5060 | "cell_type": "code", |
||
| 5061 | "execution_count": null, |
||
| 5062 | "metadata": {}, |
||
| 5063 | "outputs": [], |
||
| 5064 | "source": [] |
||
| 5065 | }, |
||
| 5066 | { |
||
| 5067 | "cell_type": "code", |
||
| 5068 | "execution_count": null, |
||
| 5069 | "metadata": {}, |
||
| 5070 | "outputs": [], |
||
| 5071 | "source": [] |
||
| 5072 | }, |
||
| 5073 | { |
||
| 5074 | "cell_type": "code", |
||
| 5075 | "execution_count": null, |
||
| 5076 | "metadata": {}, |
||
| 5077 | "outputs": [], |
||
| 5078 | "source": [] |
||
| 5079 | }, |
||
| 5080 | { |
||
| 5081 | "cell_type": "code", |
||
| 5082 | "execution_count": null, |
||
| 5083 | "metadata": {}, |
||
| 5084 | "outputs": [], |
||
| 5085 | "source": [] |
||
| 5086 | }, |
||
| 5087 | { |
||
| 5088 | "cell_type": "code", |
||
| 5089 | "execution_count": null, |
||
| 5090 | "metadata": {}, |
||
| 5091 | "outputs": [], |
||
| 5092 | "source": [] |
||
| 5093 | }, |
||
| 5094 | { |
||
| 5095 | "cell_type": "code", |
||
| 5096 | "execution_count": 104, |
||
| 5097 | "metadata": {}, |
||
| 5098 | "outputs": [ |
||
| 5099 | { |
||
| 5100 | "ename": "ModuleNotFoundError", |
||
| 5101 | "evalue": "No module named 'torch'", |
||
| 5102 | "output_type": "error", |
||
| 5103 | "traceback": [ |
||
| 5104 | "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", |
||
| 5105 | "\u001b[0;31mModuleNotFoundError\u001b[0m Traceback (most recent call last)", |
||
| 5106 | "\u001b[0;32m<ipython-input-104-6bb47b25d46a>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmath\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 2\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0;32mimport\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 4\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mtorch\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mnn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n", |
||
| 5107 | "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'torch'" |
||
| 5108 | ] |
||
| 5109 | } |
||
| 5110 | ], |
||
| 5111 | "source": [ |
||
| 5112 | "import math\n", |
||
| 5113 | "\n", |
||
| 5114 | "import torch\n", |
||
| 5115 | "from torch import nn\n", |
||
| 5116 | "\n", |
||
| 5117 | "############\n", |
||
| 5118 | "# Omic Model\n", |
||
| 5119 | "############\n", |
||
| 5120 | "def init_max_weights(module):\n", |
||
| 5121 | " for m in module.modules():\n", |
||
| 5122 | " if type(m) == nn.Linear:\n", |
||
| 5123 | " stdv = 1. / math.sqrt(m.weight.size(1))\n", |
||
| 5124 | " m.weight.data.normal_(0, stdv)\n", |
||
| 5125 | " m.bias.data.zero_()\n", |
||
| 5126 | "\n", |
||
| 5127 | "def SNN_Block(dim1, dim2, dropout=0.25):\n", |
||
| 5128 | " return nn.Sequential(\n", |
||
| 5129 | " nn.Linear(dim1, dim2),\n", |
||
| 5130 | " nn.ELU(),\n", |
||
| 5131 | " nn.AlphaDropout(p=dropout, inplace=False))\n", |
||
| 5132 | "\n", |
||
| 5133 | "class MaxNet(nn.Module):\n", |
||
| 5134 | " def __init__(self, input_dim: int, meta_dim: int=0, model_size_omic: str='small', n_classes: int=4):\n", |
||
| 5135 | " super(MaxNet, self).__init__()\n", |
||
| 5136 | " self.meta_dim = meta_dim\n", |
||
| 5137 | " self.n_classes = n_classes\n", |
||
| 5138 | " self.size_dict_omic = {'small': [256, 256, 256, 256], 'big': [1024, 1024, 1024, 256]}\n", |
||
| 5139 | " \n", |
||
| 5140 | " ### Constructing Genomic SNN\n", |
||
| 5141 | " hidden = self.size_dict_omic[model_size_omic]\n", |
||
| 5142 | " fc_omic = [SNN_Block(dim1=input_dim, dim2=hidden[0])]\n", |
||
| 5143 | " for i, _ in enumerate(hidden[1:]):\n", |
||
| 5144 | " fc_omic.append(SNN_Block(dim1=hidden[i], dim2=hidden[i+1], dropout=0.25))\n", |
||
| 5145 | " self.fc_omic = nn.Sequential(*fc_omic)\n", |
||
| 5146 | " self.classifier = nn.Linear(hidden[-1]+self.meta_dim, n_classes)\n", |
||
| 5147 | " init_max_weights(self)\n", |
||
| 5148 | "\n", |
||
| 5149 | " def forward(self, **kwargs):\n", |
||
| 5150 | " x = kwargs['x_omic']\n", |
||
| 5151 | " meta = kwargs['meta']\n", |
||
| 5152 | " features = self.fc_omic(x)\n", |
||
| 5153 | "\n", |
||
| 5154 | " if self.meta_dim: \n", |
||
| 5155 | " axis_dim = 1 if len(meta.shape) > 1 else 0\n", |
||
| 5156 | " features = torch.cat((features, meta), axis_dim)\n", |
||
| 5157 | "\n", |
||
| 5158 | " logits = self.classifier(features).unsqueeze(0)\n", |
||
| 5159 | " Y_hat = torch.topk(logits, 1, dim=1)[1]\n", |
||
| 5160 | " hazards = torch.sigmoid(logits)\n", |
||
| 5161 | " S = torch.cumprod(1 - hazards, dim=1)\n", |
||
| 5162 | " return hazards, S, Y_hat, None, None\n", |
||
| 5163 | "\n", |
||
| 5164 | " def relocate(self):\n", |
||
| 5165 | " device=torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n", |
||
| 5166 | "\n", |
||
| 5167 | " if torch.cuda.device_count() > 1:\n", |
||
| 5168 | " device_ids = list(range(torch.cuda.device_count()))\n", |
||
| 5169 | " self.fc_omic = nn.DataParallel(self.fc_omic, device_ids=device_ids).to('cuda:0')\n", |
||
| 5170 | " else:\n", |
||
| 5171 | " self.fc_omic = self.fc_omic.to(device)\n", |
||
| 5172 | "\n", |
||
| 5173 | "\n", |
||
| 5174 | " self.classifier = self.classifier.to(device)" |
||
| 5175 | ] |
||
| 5176 | }, |
||
| 5177 | { |
||
| 5178 | "cell_type": "code", |
||
| 5179 | "execution_count": null, |
||
| 5180 | "metadata": {}, |
||
| 5181 | "outputs": [], |
||
| 5182 | "source": [] |
||
| 5183 | }, |
||
| 5184 | { |
||
| 5185 | "cell_type": "code", |
||
| 5186 | "execution_count": 88, |
||
| 5187 | "metadata": {}, |
||
| 5188 | "outputs": [ |
||
| 5189 | { |
||
| 5190 | "data": { |
||
| 5191 | "text/html": [ |
||
| 5192 | "<div>\n", |
||
| 5193 | "<style scoped>\n", |
||
| 5194 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 5195 | " vertical-align: middle;\n", |
||
| 5196 | " }\n", |
||
| 5197 | "\n", |
||
| 5198 | " .dataframe tbody tr th {\n", |
||
| 5199 | " vertical-align: top;\n", |
||
| 5200 | " }\n", |
||
| 5201 | "\n", |
||
| 5202 | " .dataframe thead th {\n", |
||
| 5203 | " text-align: right;\n", |
||
| 5204 | " }\n", |
||
| 5205 | "</style>\n", |
||
| 5206 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 5207 | " <thead>\n", |
||
| 5208 | " <tr style=\"text-align: right;\">\n", |
||
| 5209 | " <th></th>\n", |
||
| 5210 | " <th>CXCL14_rnaseq</th>\n", |
||
| 5211 | " <th>FGF1_rnaseq</th>\n", |
||
| 5212 | " <th>IFNA8_cnv</th>\n", |
||
| 5213 | " <th>ADM_rnaseq</th>\n", |
||
| 5214 | " <th>LTBP2_rnaseq</th>\n", |
||
| 5215 | " <th>CCL28_rnaseq</th>\n", |
||
| 5216 | " <th>IFNA7_rnaseq</th>\n", |
||
| 5217 | " <th>GH2_rnaseq</th>\n", |
||
| 5218 | " <th>AIMP1_rnaseq</th>\n", |
||
| 5219 | " <th>DEFB1_rnaseq</th>\n", |
||
| 5220 | " <th>...</th>\n", |
||
| 5221 | " <th>NPPB_rnaseq</th>\n", |
||
| 5222 | " <th>CCL27_rnaseq</th>\n", |
||
| 5223 | " <th>FASLG_rnaseq</th>\n", |
||
| 5224 | " <th>FGF20_cnv</th>\n", |
||
| 5225 | " <th>FAM3C_rnaseq</th>\n", |
||
| 5226 | " <th>IL18_rnaseq</th>\n", |
||
| 5227 | " <th>GDF10_rnaseq</th>\n", |
||
| 5228 | " <th>MYDGF_rnaseq</th>\n", |
||
| 5229 | " <th>IL10_rnaseq</th>\n", |
||
| 5230 | " <th>IFNW1_rnaseq</th>\n", |
||
| 5231 | " </tr>\n", |
||
| 5232 | " </thead>\n", |
||
| 5233 | " <tbody>\n", |
||
| 5234 | " <tr>\n", |
||
| 5235 | " <th>0</th>\n", |
||
| 5236 | " <td>-0.1170</td>\n", |
||
| 5237 | " <td>-0.2221</td>\n", |
||
| 5238 | " <td>1</td>\n", |
||
| 5239 | " <td>-0.5126</td>\n", |
||
| 5240 | " <td>-0.3289</td>\n", |
||
| 5241 | " <td>-0.7331</td>\n", |
||
| 5242 | " <td>-0.1244</td>\n", |
||
| 5243 | " <td>-0.1693</td>\n", |
||
| 5244 | " <td>0.5942</td>\n", |
||
| 5245 | " <td>-0.4707</td>\n", |
||
| 5246 | " <td>...</td>\n", |
||
| 5247 | " <td>-0.2276</td>\n", |
||
| 5248 | " <td>1.2033</td>\n", |
||
| 5249 | " <td>0.9826</td>\n", |
||
| 5250 | " <td>-1</td>\n", |
||
| 5251 | " <td>-0.6161</td>\n", |
||
| 5252 | " <td>-0.5643</td>\n", |
||
| 5253 | " <td>-0.2165</td>\n", |
||
| 5254 | " <td>-0.2836</td>\n", |
||
| 5255 | " <td>0.9991</td>\n", |
||
| 5256 | " <td>-0.3899</td>\n", |
||
| 5257 | " </tr>\n", |
||
| 5258 | " <tr>\n", |
||
| 5259 | " <th>1</th>\n", |
||
| 5260 | " <td>-0.2330</td>\n", |
||
| 5261 | " <td>-0.4343</td>\n", |
||
| 5262 | " <td>-1</td>\n", |
||
| 5263 | " <td>-0.2381</td>\n", |
||
| 5264 | " <td>-0.4799</td>\n", |
||
| 5265 | " <td>-0.0520</td>\n", |
||
| 5266 | " <td>-0.1244</td>\n", |
||
| 5267 | " <td>-0.1693</td>\n", |
||
| 5268 | " <td>1.1854</td>\n", |
||
| 5269 | " <td>-0.4820</td>\n", |
||
| 5270 | " <td>...</td>\n", |
||
| 5271 | " <td>-0.2276</td>\n", |
||
| 5272 | " <td>-0.2946</td>\n", |
||
| 5273 | " <td>-0.5443</td>\n", |
||
| 5274 | " <td>-1</td>\n", |
||
| 5275 | " <td>-0.3499</td>\n", |
||
| 5276 | " <td>-0.7958</td>\n", |
||
| 5277 | " <td>-0.3140</td>\n", |
||
| 5278 | " <td>-0.3359</td>\n", |
||
| 5279 | " <td>-0.4865</td>\n", |
||
| 5280 | " <td>-0.3899</td>\n", |
||
| 5281 | " </tr>\n", |
||
| 5282 | " <tr>\n", |
||
| 5283 | " <th>2</th>\n", |
||
| 5284 | " <td>-0.1384</td>\n", |
||
| 5285 | " <td>-0.1597</td>\n", |
||
| 5286 | " <td>-1</td>\n", |
||
| 5287 | " <td>-0.1521</td>\n", |
||
| 5288 | " <td>-0.3348</td>\n", |
||
| 5289 | " <td>-0.5310</td>\n", |
||
| 5290 | " <td>-0.1244</td>\n", |
||
| 5291 | " <td>-0.1693</td>\n", |
||
| 5292 | " <td>0.3889</td>\n", |
||
| 5293 | " <td>-0.3607</td>\n", |
||
| 5294 | " <td>...</td>\n", |
||
| 5295 | " <td>3.4177</td>\n", |
||
| 5296 | " <td>-0.2946</td>\n", |
||
| 5297 | " <td>-0.5320</td>\n", |
||
| 5298 | " <td>0</td>\n", |
||
| 5299 | " <td>0.4581</td>\n", |
||
| 5300 | " <td>-0.6179</td>\n", |
||
| 5301 | " <td>-0.2107</td>\n", |
||
| 5302 | " <td>0.2751</td>\n", |
||
| 5303 | " <td>-0.5108</td>\n", |
||
| 5304 | " <td>1.0629</td>\n", |
||
| 5305 | " </tr>\n", |
||
| 5306 | " <tr>\n", |
||
| 5307 | " <th>3</th>\n", |
||
| 5308 | " <td>-0.1624</td>\n", |
||
| 5309 | " <td>-0.3463</td>\n", |
||
| 5310 | " <td>-1</td>\n", |
||
| 5311 | " <td>0.0272</td>\n", |
||
| 5312 | " <td>-0.7623</td>\n", |
||
| 5313 | " <td>0.8196</td>\n", |
||
| 5314 | " <td>-0.1244</td>\n", |
||
| 5315 | " <td>-0.1693</td>\n", |
||
| 5316 | " <td>-0.0416</td>\n", |
||
| 5317 | " <td>0.1661</td>\n", |
||
| 5318 | " <td>...</td>\n", |
||
| 5319 | " <td>-0.2276</td>\n", |
||
| 5320 | " <td>-0.1020</td>\n", |
||
| 5321 | " <td>-0.4682</td>\n", |
||
| 5322 | " <td>-1</td>\n", |
||
| 5323 | " <td>-0.4391</td>\n", |
||
| 5324 | " <td>-0.7275</td>\n", |
||
| 5325 | " <td>-0.2876</td>\n", |
||
| 5326 | " <td>-0.4696</td>\n", |
||
| 5327 | " <td>-0.6248</td>\n", |
||
| 5328 | " <td>-0.3899</td>\n", |
||
| 5329 | " </tr>\n", |
||
| 5330 | " <tr>\n", |
||
| 5331 | " <th>4</th>\n", |
||
| 5332 | " <td>-0.2346</td>\n", |
||
| 5333 | " <td>-0.4090</td>\n", |
||
| 5334 | " <td>-1</td>\n", |
||
| 5335 | " <td>-0.2078</td>\n", |
||
| 5336 | " <td>0.5702</td>\n", |
||
| 5337 | " <td>-0.4219</td>\n", |
||
| 5338 | " <td>-0.1244</td>\n", |
||
| 5339 | " <td>0.5257</td>\n", |
||
| 5340 | " <td>-0.9790</td>\n", |
||
| 5341 | " <td>0.3938</td>\n", |
||
| 5342 | " <td>...</td>\n", |
||
| 5343 | " <td>-0.2276</td>\n", |
||
| 5344 | " <td>-0.1035</td>\n", |
||
| 5345 | " <td>-0.4688</td>\n", |
||
| 5346 | " <td>-1</td>\n", |
||
| 5347 | " <td>1.2596</td>\n", |
||
| 5348 | " <td>-0.5807</td>\n", |
||
| 5349 | " <td>0.4108</td>\n", |
||
| 5350 | " <td>0.1801</td>\n", |
||
| 5351 | " <td>-0.6086</td>\n", |
||
| 5352 | " <td>-0.3899</td>\n", |
||
| 5353 | " </tr>\n", |
||
| 5354 | " <tr>\n", |
||
| 5355 | " <th>...</th>\n", |
||
| 5356 | " <td>...</td>\n", |
||
| 5357 | " <td>...</td>\n", |
||
| 5358 | " <td>...</td>\n", |
||
| 5359 | " <td>...</td>\n", |
||
| 5360 | " <td>...</td>\n", |
||
| 5361 | " <td>...</td>\n", |
||
| 5362 | " <td>...</td>\n", |
||
| 5363 | " <td>...</td>\n", |
||
| 5364 | " <td>...</td>\n", |
||
| 5365 | " <td>...</td>\n", |
||
| 5366 | " <td>...</td>\n", |
||
| 5367 | " <td>...</td>\n", |
||
| 5368 | " <td>...</td>\n", |
||
| 5369 | " <td>...</td>\n", |
||
| 5370 | " <td>...</td>\n", |
||
| 5371 | " <td>...</td>\n", |
||
| 5372 | " <td>...</td>\n", |
||
| 5373 | " <td>...</td>\n", |
||
| 5374 | " <td>...</td>\n", |
||
| 5375 | " <td>...</td>\n", |
||
| 5376 | " <td>...</td>\n", |
||
| 5377 | " </tr>\n", |
||
| 5378 | " <tr>\n", |
||
| 5379 | " <th>368</th>\n", |
||
| 5380 | " <td>-0.2417</td>\n", |
||
| 5381 | " <td>10.1423</td>\n", |
||
| 5382 | " <td>-1</td>\n", |
||
| 5383 | " <td>-0.5456</td>\n", |
||
| 5384 | " <td>0.8742</td>\n", |
||
| 5385 | " <td>-0.1822</td>\n", |
||
| 5386 | " <td>-0.1244</td>\n", |
||
| 5387 | " <td>-0.1693</td>\n", |
||
| 5388 | " <td>-1.2395</td>\n", |
||
| 5389 | " <td>-0.5125</td>\n", |
||
| 5390 | " <td>...</td>\n", |
||
| 5391 | " <td>-0.2276</td>\n", |
||
| 5392 | " <td>-0.2946</td>\n", |
||
| 5393 | " <td>0.0777</td>\n", |
||
| 5394 | " <td>0</td>\n", |
||
| 5395 | " <td>-0.8242</td>\n", |
||
| 5396 | " <td>-0.6727</td>\n", |
||
| 5397 | " <td>0.1938</td>\n", |
||
| 5398 | " <td>0.9210</td>\n", |
||
| 5399 | " <td>0.4479</td>\n", |
||
| 5400 | " <td>-0.3899</td>\n", |
||
| 5401 | " </tr>\n", |
||
| 5402 | " <tr>\n", |
||
| 5403 | " <th>369</th>\n", |
||
| 5404 | " <td>-0.2412</td>\n", |
||
| 5405 | " <td>1.3253</td>\n", |
||
| 5406 | " <td>1</td>\n", |
||
| 5407 | " <td>-0.5680</td>\n", |
||
| 5408 | " <td>1.0719</td>\n", |
||
| 5409 | " <td>-0.1707</td>\n", |
||
| 5410 | " <td>-0.1244</td>\n", |
||
| 5411 | " <td>-0.1693</td>\n", |
||
| 5412 | " <td>-1.6694</td>\n", |
||
| 5413 | " <td>-0.4528</td>\n", |
||
| 5414 | " <td>...</td>\n", |
||
| 5415 | " <td>0.5679</td>\n", |
||
| 5416 | " <td>-0.2661</td>\n", |
||
| 5417 | " <td>1.0215</td>\n", |
||
| 5418 | " <td>-2</td>\n", |
||
| 5419 | " <td>-0.5327</td>\n", |
||
| 5420 | " <td>0.3335</td>\n", |
||
| 5421 | " <td>-0.1730</td>\n", |
||
| 5422 | " <td>0.0147</td>\n", |
||
| 5423 | " <td>0.6012</td>\n", |
||
| 5424 | " <td>2.2526</td>\n", |
||
| 5425 | " </tr>\n", |
||
| 5426 | " <tr>\n", |
||
| 5427 | " <th>370</th>\n", |
||
| 5428 | " <td>-0.2396</td>\n", |
||
| 5429 | " <td>0.0435</td>\n", |
||
| 5430 | " <td>0</td>\n", |
||
| 5431 | " <td>-0.3610</td>\n", |
||
| 5432 | " <td>3.1965</td>\n", |
||
| 5433 | " <td>1.3670</td>\n", |
||
| 5434 | " <td>-0.1244</td>\n", |
||
| 5435 | " <td>-0.1693</td>\n", |
||
| 5436 | " <td>0.4439</td>\n", |
||
| 5437 | " <td>-0.5099</td>\n", |
||
| 5438 | " <td>...</td>\n", |
||
| 5439 | " <td>-0.2276</td>\n", |
||
| 5440 | " <td>-0.2289</td>\n", |
||
| 5441 | " <td>0.0521</td>\n", |
||
| 5442 | " <td>-1</td>\n", |
||
| 5443 | " <td>1.0317</td>\n", |
||
| 5444 | " <td>-0.1473</td>\n", |
||
| 5445 | " <td>-0.1517</td>\n", |
||
| 5446 | " <td>0.9384</td>\n", |
||
| 5447 | " <td>-0.3165</td>\n", |
||
| 5448 | " <td>0.6239</td>\n", |
||
| 5449 | " </tr>\n", |
||
| 5450 | " <tr>\n", |
||
| 5451 | " <th>371</th>\n", |
||
| 5452 | " <td>-0.2393</td>\n", |
||
| 5453 | " <td>-0.4475</td>\n", |
||
| 5454 | " <td>0</td>\n", |
||
| 5455 | " <td>0.4772</td>\n", |
||
| 5456 | " <td>2.9612</td>\n", |
||
| 5457 | " <td>-0.7799</td>\n", |
||
| 5458 | " <td>-0.1244</td>\n", |
||
| 5459 | " <td>-0.1693</td>\n", |
||
| 5460 | " <td>0.5778</td>\n", |
||
| 5461 | " <td>1.7607</td>\n", |
||
| 5462 | " <td>...</td>\n", |
||
| 5463 | " <td>-0.2276</td>\n", |
||
| 5464 | " <td>9.4098</td>\n", |
||
| 5465 | " <td>-0.5443</td>\n", |
||
| 5466 | " <td>0</td>\n", |
||
| 5467 | " <td>0.2992</td>\n", |
||
| 5468 | " <td>-0.5451</td>\n", |
||
| 5469 | " <td>-0.2456</td>\n", |
||
| 5470 | " <td>0.8898</td>\n", |
||
| 5471 | " <td>-0.5781</td>\n", |
||
| 5472 | " <td>-0.3899</td>\n", |
||
| 5473 | " </tr>\n", |
||
| 5474 | " <tr>\n", |
||
| 5475 | " <th>372</th>\n", |
||
| 5476 | " <td>-0.1936</td>\n", |
||
| 5477 | " <td>-0.2281</td>\n", |
||
| 5478 | " <td>0</td>\n", |
||
| 5479 | " <td>-0.4124</td>\n", |
||
| 5480 | " <td>-0.1873</td>\n", |
||
| 5481 | " <td>-0.1200</td>\n", |
||
| 5482 | " <td>-0.1244</td>\n", |
||
| 5483 | " <td>-0.0326</td>\n", |
||
| 5484 | " <td>-0.8786</td>\n", |
||
| 5485 | " <td>-0.3912</td>\n", |
||
| 5486 | " <td>...</td>\n", |
||
| 5487 | " <td>-0.2276</td>\n", |
||
| 5488 | " <td>-0.2570</td>\n", |
||
| 5489 | " <td>-0.3810</td>\n", |
||
| 5490 | " <td>-1</td>\n", |
||
| 5491 | " <td>-0.6399</td>\n", |
||
| 5492 | " <td>-0.9128</td>\n", |
||
| 5493 | " <td>0.3367</td>\n", |
||
| 5494 | " <td>-0.4686</td>\n", |
||
| 5495 | " <td>0.8995</td>\n", |
||
| 5496 | " <td>1.3522</td>\n", |
||
| 5497 | " </tr>\n", |
||
| 5498 | " </tbody>\n", |
||
| 5499 | "</table>\n", |
||
| 5500 | "<p>373 rows × 347 columns</p>\n", |
||
| 5501 | "</div>" |
||
| 5502 | ], |
||
| 5503 | "text/plain": [ |
||
| 5504 | " CXCL14_rnaseq FGF1_rnaseq IFNA8_cnv ADM_rnaseq LTBP2_rnaseq \\\n", |
||
| 5505 | "0 -0.1170 -0.2221 1 -0.5126 -0.3289 \n", |
||
| 5506 | "1 -0.2330 -0.4343 -1 -0.2381 -0.4799 \n", |
||
| 5507 | "2 -0.1384 -0.1597 -1 -0.1521 -0.3348 \n", |
||
| 5508 | "3 -0.1624 -0.3463 -1 0.0272 -0.7623 \n", |
||
| 5509 | "4 -0.2346 -0.4090 -1 -0.2078 0.5702 \n", |
||
| 5510 | ".. ... ... ... ... ... \n", |
||
| 5511 | "368 -0.2417 10.1423 -1 -0.5456 0.8742 \n", |
||
| 5512 | "369 -0.2412 1.3253 1 -0.5680 1.0719 \n", |
||
| 5513 | "370 -0.2396 0.0435 0 -0.3610 3.1965 \n", |
||
| 5514 | "371 -0.2393 -0.4475 0 0.4772 2.9612 \n", |
||
| 5515 | "372 -0.1936 -0.2281 0 -0.4124 -0.1873 \n", |
||
| 5516 | "\n", |
||
| 5517 | " CCL28_rnaseq IFNA7_rnaseq GH2_rnaseq AIMP1_rnaseq DEFB1_rnaseq ... \\\n", |
||
| 5518 | "0 -0.7331 -0.1244 -0.1693 0.5942 -0.4707 ... \n", |
||
| 5519 | "1 -0.0520 -0.1244 -0.1693 1.1854 -0.4820 ... \n", |
||
| 5520 | "2 -0.5310 -0.1244 -0.1693 0.3889 -0.3607 ... \n", |
||
| 5521 | "3 0.8196 -0.1244 -0.1693 -0.0416 0.1661 ... \n", |
||
| 5522 | "4 -0.4219 -0.1244 0.5257 -0.9790 0.3938 ... \n", |
||
| 5523 | ".. ... ... ... ... ... ... \n", |
||
| 5524 | "368 -0.1822 -0.1244 -0.1693 -1.2395 -0.5125 ... \n", |
||
| 5525 | "369 -0.1707 -0.1244 -0.1693 -1.6694 -0.4528 ... \n", |
||
| 5526 | "370 1.3670 -0.1244 -0.1693 0.4439 -0.5099 ... \n", |
||
| 5527 | "371 -0.7799 -0.1244 -0.1693 0.5778 1.7607 ... \n", |
||
| 5528 | "372 -0.1200 -0.1244 -0.0326 -0.8786 -0.3912 ... \n", |
||
| 5529 | "\n", |
||
| 5530 | " NPPB_rnaseq CCL27_rnaseq FASLG_rnaseq FGF20_cnv FAM3C_rnaseq \\\n", |
||
| 5531 | "0 -0.2276 1.2033 0.9826 -1 -0.6161 \n", |
||
| 5532 | "1 -0.2276 -0.2946 -0.5443 -1 -0.3499 \n", |
||
| 5533 | "2 3.4177 -0.2946 -0.5320 0 0.4581 \n", |
||
| 5534 | "3 -0.2276 -0.1020 -0.4682 -1 -0.4391 \n", |
||
| 5535 | "4 -0.2276 -0.1035 -0.4688 -1 1.2596 \n", |
||
| 5536 | ".. ... ... ... ... ... \n", |
||
| 5537 | "368 -0.2276 -0.2946 0.0777 0 -0.8242 \n", |
||
| 5538 | "369 0.5679 -0.2661 1.0215 -2 -0.5327 \n", |
||
| 5539 | "370 -0.2276 -0.2289 0.0521 -1 1.0317 \n", |
||
| 5540 | "371 -0.2276 9.4098 -0.5443 0 0.2992 \n", |
||
| 5541 | "372 -0.2276 -0.2570 -0.3810 -1 -0.6399 \n", |
||
| 5542 | "\n", |
||
| 5543 | " IL18_rnaseq GDF10_rnaseq MYDGF_rnaseq IL10_rnaseq IFNW1_rnaseq \n", |
||
| 5544 | "0 -0.5643 -0.2165 -0.2836 0.9991 -0.3899 \n", |
||
| 5545 | "1 -0.7958 -0.3140 -0.3359 -0.4865 -0.3899 \n", |
||
| 5546 | "2 -0.6179 -0.2107 0.2751 -0.5108 1.0629 \n", |
||
| 5547 | "3 -0.7275 -0.2876 -0.4696 -0.6248 -0.3899 \n", |
||
| 5548 | "4 -0.5807 0.4108 0.1801 -0.6086 -0.3899 \n", |
||
| 5549 | ".. ... ... ... ... ... \n", |
||
| 5550 | "368 -0.6727 0.1938 0.9210 0.4479 -0.3899 \n", |
||
| 5551 | "369 0.3335 -0.1730 0.0147 0.6012 2.2526 \n", |
||
| 5552 | "370 -0.1473 -0.1517 0.9384 -0.3165 0.6239 \n", |
||
| 5553 | "371 -0.5451 -0.2456 0.8898 -0.5781 -0.3899 \n", |
||
| 5554 | "372 -0.9128 0.3367 -0.4686 0.8995 1.3522 \n", |
||
| 5555 | "\n", |
||
| 5556 | "[373 rows x 347 columns]" |
||
| 5557 | ] |
||
| 5558 | }, |
||
| 5559 | "execution_count": 88, |
||
| 5560 | "metadata": {}, |
||
| 5561 | "output_type": "execute_result" |
||
| 5562 | } |
||
| 5563 | ], |
||
| 5564 | "source": [ |
||
| 5565 | "genomic_features[series_intersecdef series_intersection(s1, s2):\n", |
||
| 5566 | " return pd.Series(list(set(s1) & set(s2)))\n", |
||
| 5567 | "tion(sig, genomic_features.columns)]" |
||
| 5568 | ] |
||
| 5569 | }, |
||
| 5570 | { |
||
| 5571 | "cell_type": "code", |
||
| 5572 | "execution_count": 84, |
||
| 5573 | "metadata": {}, |
||
| 5574 | "outputs": [], |
||
| 5575 | "source": [ |
||
| 5576 | "def series_intersection(s1, s2):\n", |
||
| 5577 | " return pd.Series(list(set(s1) & set(s2)))\n" |
||
| 5578 | ] |
||
| 5579 | }, |
||
| 5580 | { |
||
| 5581 | "cell_type": "code", |
||
| 5582 | "execution_count": 68, |
||
| 5583 | "metadata": {}, |
||
| 5584 | "outputs": [ |
||
| 5585 | { |
||
| 5586 | "data": { |
||
| 5587 | "text/html": [ |
||
| 5588 | "<div>\n", |
||
| 5589 | "<style scoped>\n", |
||
| 5590 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 5591 | " vertical-align: middle;\n", |
||
| 5592 | " }\n", |
||
| 5593 | "\n", |
||
| 5594 | " .dataframe tbody tr th {\n", |
||
| 5595 | " vertical-align: top;\n", |
||
| 5596 | " }\n", |
||
| 5597 | "\n", |
||
| 5598 | " .dataframe thead th {\n", |
||
| 5599 | " text-align: right;\n", |
||
| 5600 | " }\n", |
||
| 5601 | "</style>\n", |
||
| 5602 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 5603 | " <thead>\n", |
||
| 5604 | " <tr style=\"text-align: right;\">\n", |
||
| 5605 | " <th></th>\n", |
||
| 5606 | " <th>NDUFS5_cnv</th>\n", |
||
| 5607 | " <th>MACF1_cnv</th>\n", |
||
| 5608 | " <th>RNA5SP44_cnv</th>\n", |
||
| 5609 | " <th>KIAA0754_cnv</th>\n", |
||
| 5610 | " <th>BMP8A_cnv</th>\n", |
||
| 5611 | " <th>PABPC4_cnv</th>\n", |
||
| 5612 | " <th>SNORA55_cnv</th>\n", |
||
| 5613 | " <th>HEYL_cnv</th>\n", |
||
| 5614 | " <th>HPCAL4_cnv</th>\n", |
||
| 5615 | " <th>NT5C1A_cnv</th>\n", |
||
| 5616 | " <th>...</th>\n", |
||
| 5617 | " <th>ZWINT_rnaseq</th>\n", |
||
| 5618 | " <th>ZXDA_rnaseq</th>\n", |
||
| 5619 | " <th>ZXDB_rnaseq</th>\n", |
||
| 5620 | " <th>ZXDC_rnaseq</th>\n", |
||
| 5621 | " <th>ZYG11A_rnaseq</th>\n", |
||
| 5622 | " <th>ZYG11B_rnaseq</th>\n", |
||
| 5623 | " <th>ZYX_rnaseq</th>\n", |
||
| 5624 | " <th>ZZEF1_rnaseq</th>\n", |
||
| 5625 | " <th>ZZZ3_rnaseq</th>\n", |
||
| 5626 | " <th>TPTEP1_rnaseq</th>\n", |
||
| 5627 | " </tr>\n", |
||
| 5628 | " </thead>\n", |
||
| 5629 | " <tbody>\n", |
||
| 5630 | " <tr>\n", |
||
| 5631 | " <th>0</th>\n", |
||
| 5632 | " <td>-1</td>\n", |
||
| 5633 | " <td>-1</td>\n", |
||
| 5634 | " <td>-1</td>\n", |
||
| 5635 | " <td>-1</td>\n", |
||
| 5636 | " <td>-1</td>\n", |
||
| 5637 | " <td>-1</td>\n", |
||
| 5638 | " <td>-1</td>\n", |
||
| 5639 | " <td>-1</td>\n", |
||
| 5640 | " <td>-1</td>\n", |
||
| 5641 | " <td>-1</td>\n", |
||
| 5642 | " <td>...</td>\n", |
||
| 5643 | " <td>-0.8388</td>\n", |
||
| 5644 | " <td>4.1375</td>\n", |
||
| 5645 | " <td>3.9664</td>\n", |
||
| 5646 | " <td>1.8437</td>\n", |
||
| 5647 | " <td>-0.3959</td>\n", |
||
| 5648 | " <td>-0.2561</td>\n", |
||
| 5649 | " <td>-0.2866</td>\n", |
||
| 5650 | " <td>1.8770</td>\n", |
||
| 5651 | " <td>-0.3179</td>\n", |
||
| 5652 | " <td>-0.3633</td>\n", |
||
| 5653 | " </tr>\n", |
||
| 5654 | " <tr>\n", |
||
| 5655 | " <th>1</th>\n", |
||
| 5656 | " <td>2</td>\n", |
||
| 5657 | " <td>2</td>\n", |
||
| 5658 | " <td>2</td>\n", |
||
| 5659 | " <td>2</td>\n", |
||
| 5660 | " <td>2</td>\n", |
||
| 5661 | " <td>2</td>\n", |
||
| 5662 | " <td>2</td>\n", |
||
| 5663 | " <td>2</td>\n", |
||
| 5664 | " <td>2</td>\n", |
||
| 5665 | " <td>2</td>\n", |
||
| 5666 | " <td>...</td>\n", |
||
| 5667 | " <td>-0.1083</td>\n", |
||
| 5668 | " <td>0.3393</td>\n", |
||
| 5669 | " <td>0.2769</td>\n", |
||
| 5670 | " <td>1.7320</td>\n", |
||
| 5671 | " <td>-0.0975</td>\n", |
||
| 5672 | " <td>2.6955</td>\n", |
||
| 5673 | " <td>-0.6741</td>\n", |
||
| 5674 | " <td>1.0323</td>\n", |
||
| 5675 | " <td>1.2766</td>\n", |
||
| 5676 | " <td>-0.3982</td>\n", |
||
| 5677 | " </tr>\n", |
||
| 5678 | " <tr>\n", |
||
| 5679 | " <th>2</th>\n", |
||
| 5680 | " <td>0</td>\n", |
||
| 5681 | " <td>0</td>\n", |
||
| 5682 | " <td>0</td>\n", |
||
| 5683 | " <td>0</td>\n", |
||
| 5684 | " <td>0</td>\n", |
||
| 5685 | " <td>0</td>\n", |
||
| 5686 | " <td>0</td>\n", |
||
| 5687 | " <td>0</td>\n", |
||
| 5688 | " <td>0</td>\n", |
||
| 5689 | " <td>0</td>\n", |
||
| 5690 | " <td>...</td>\n", |
||
| 5691 | " <td>-0.4155</td>\n", |
||
| 5692 | " <td>1.6846</td>\n", |
||
| 5693 | " <td>0.7711</td>\n", |
||
| 5694 | " <td>-0.3061</td>\n", |
||
| 5695 | " <td>-0.5016</td>\n", |
||
| 5696 | " <td>2.8548</td>\n", |
||
| 5697 | " <td>-0.6171</td>\n", |
||
| 5698 | " <td>-0.8608</td>\n", |
||
| 5699 | " <td>-0.0486</td>\n", |
||
| 5700 | " <td>-0.3962</td>\n", |
||
| 5701 | " </tr>\n", |
||
| 5702 | " <tr>\n", |
||
| 5703 | " <th>3</th>\n", |
||
| 5704 | " <td>0</td>\n", |
||
| 5705 | " <td>0</td>\n", |
||
| 5706 | " <td>0</td>\n", |
||
| 5707 | " <td>0</td>\n", |
||
| 5708 | " <td>0</td>\n", |
||
| 5709 | " <td>0</td>\n", |
||
| 5710 | " <td>0</td>\n", |
||
| 5711 | " <td>0</td>\n", |
||
| 5712 | " <td>0</td>\n", |
||
| 5713 | " <td>0</td>\n", |
||
| 5714 | " <td>...</td>\n", |
||
| 5715 | " <td>-0.8143</td>\n", |
||
| 5716 | " <td>0.8344</td>\n", |
||
| 5717 | " <td>1.5075</td>\n", |
||
| 5718 | " <td>3.6068</td>\n", |
||
| 5719 | " <td>-0.5004</td>\n", |
||
| 5720 | " <td>-0.0747</td>\n", |
||
| 5721 | " <td>-0.2185</td>\n", |
||
| 5722 | " <td>-0.4379</td>\n", |
||
| 5723 | " <td>1.6913</td>\n", |
||
| 5724 | " <td>1.7748</td>\n", |
||
| 5725 | " </tr>\n", |
||
| 5726 | " <tr>\n", |
||
| 5727 | " <th>4</th>\n", |
||
| 5728 | " <td>0</td>\n", |
||
| 5729 | " <td>0</td>\n", |
||
| 5730 | " <td>0</td>\n", |
||
| 5731 | " <td>0</td>\n", |
||
| 5732 | " <td>0</td>\n", |
||
| 5733 | " <td>0</td>\n", |
||
| 5734 | " <td>0</td>\n", |
||
| 5735 | " <td>0</td>\n", |
||
| 5736 | " <td>0</td>\n", |
||
| 5737 | " <td>0</td>\n", |
||
| 5738 | " <td>...</td>\n", |
||
| 5739 | " <td>0.0983</td>\n", |
||
| 5740 | " <td>-0.7908</td>\n", |
||
| 5741 | " <td>-0.0053</td>\n", |
||
| 5742 | " <td>-0.0643</td>\n", |
||
| 5743 | " <td>-0.3706</td>\n", |
||
| 5744 | " <td>0.3870</td>\n", |
||
| 5745 | " <td>-0.5589</td>\n", |
||
| 5746 | " <td>-0.5979</td>\n", |
||
| 5747 | " <td>0.0047</td>\n", |
||
| 5748 | " <td>-0.3548</td>\n", |
||
| 5749 | " </tr>\n", |
||
| 5750 | " <tr>\n", |
||
| 5751 | " <th>...</th>\n", |
||
| 5752 | " <td>...</td>\n", |
||
| 5753 | " <td>...</td>\n", |
||
| 5754 | " <td>...</td>\n", |
||
| 5755 | " <td>...</td>\n", |
||
| 5756 | " <td>...</td>\n", |
||
| 5757 | " <td>...</td>\n", |
||
| 5758 | " <td>...</td>\n", |
||
| 5759 | " <td>...</td>\n", |
||
| 5760 | " <td>...</td>\n", |
||
| 5761 | " <td>...</td>\n", |
||
| 5762 | " <td>...</td>\n", |
||
| 5763 | " <td>...</td>\n", |
||
| 5764 | " <td>...</td>\n", |
||
| 5765 | " <td>...</td>\n", |
||
| 5766 | " <td>...</td>\n", |
||
| 5767 | " <td>...</td>\n", |
||
| 5768 | " <td>...</td>\n", |
||
| 5769 | " <td>...</td>\n", |
||
| 5770 | " <td>...</td>\n", |
||
| 5771 | " <td>...</td>\n", |
||
| 5772 | " <td>...</td>\n", |
||
| 5773 | " </tr>\n", |
||
| 5774 | " <tr>\n", |
||
| 5775 | " <th>368</th>\n", |
||
| 5776 | " <td>2</td>\n", |
||
| 5777 | " <td>2</td>\n", |
||
| 5778 | " <td>2</td>\n", |
||
| 5779 | " <td>2</td>\n", |
||
| 5780 | " <td>2</td>\n", |
||
| 5781 | " <td>2</td>\n", |
||
| 5782 | " <td>2</td>\n", |
||
| 5783 | " <td>2</td>\n", |
||
| 5784 | " <td>2</td>\n", |
||
| 5785 | " <td>2</td>\n", |
||
| 5786 | " <td>...</td>\n", |
||
| 5787 | " <td>-0.0291</td>\n", |
||
| 5788 | " <td>-0.1058</td>\n", |
||
| 5789 | " <td>-0.6721</td>\n", |
||
| 5790 | " <td>0.2802</td>\n", |
||
| 5791 | " <td>1.9504</td>\n", |
||
| 5792 | " <td>-0.8784</td>\n", |
||
| 5793 | " <td>0.9506</td>\n", |
||
| 5794 | " <td>0.0607</td>\n", |
||
| 5795 | " <td>1.1883</td>\n", |
||
| 5796 | " <td>-0.3521</td>\n", |
||
| 5797 | " </tr>\n", |
||
| 5798 | " <tr>\n", |
||
| 5799 | " <th>369</th>\n", |
||
| 5800 | " <td>0</td>\n", |
||
| 5801 | " <td>0</td>\n", |
||
| 5802 | " <td>0</td>\n", |
||
| 5803 | " <td>0</td>\n", |
||
| 5804 | " <td>0</td>\n", |
||
| 5805 | " <td>0</td>\n", |
||
| 5806 | " <td>0</td>\n", |
||
| 5807 | " <td>0</td>\n", |
||
| 5808 | " <td>0</td>\n", |
||
| 5809 | " <td>0</td>\n", |
||
| 5810 | " <td>...</td>\n", |
||
| 5811 | " <td>0.0497</td>\n", |
||
| 5812 | " <td>0.3673</td>\n", |
||
| 5813 | " <td>-0.2208</td>\n", |
||
| 5814 | " <td>0.3034</td>\n", |
||
| 5815 | " <td>3.2580</td>\n", |
||
| 5816 | " <td>-0.2089</td>\n", |
||
| 5817 | " <td>1.6053</td>\n", |
||
| 5818 | " <td>-0.8746</td>\n", |
||
| 5819 | " <td>-0.4491</td>\n", |
||
| 5820 | " <td>-0.3450</td>\n", |
||
| 5821 | " </tr>\n", |
||
| 5822 | " <tr>\n", |
||
| 5823 | " <th>370</th>\n", |
||
| 5824 | " <td>1</td>\n", |
||
| 5825 | " <td>1</td>\n", |
||
| 5826 | " <td>1</td>\n", |
||
| 5827 | " <td>1</td>\n", |
||
| 5828 | " <td>1</td>\n", |
||
| 5829 | " <td>1</td>\n", |
||
| 5830 | " <td>1</td>\n", |
||
| 5831 | " <td>1</td>\n", |
||
| 5832 | " <td>1</td>\n", |
||
| 5833 | " <td>1</td>\n", |
||
| 5834 | " <td>...</td>\n", |
||
| 5835 | " <td>0.3822</td>\n", |
||
| 5836 | " <td>-0.7003</td>\n", |
||
| 5837 | " <td>-0.7661</td>\n", |
||
| 5838 | " <td>-1.7035</td>\n", |
||
| 5839 | " <td>-0.5423</td>\n", |
||
| 5840 | " <td>-0.3488</td>\n", |
||
| 5841 | " <td>1.3713</td>\n", |
||
| 5842 | " <td>-0.4365</td>\n", |
||
| 5843 | " <td>2.3456</td>\n", |
||
| 5844 | " <td>-0.3866</td>\n", |
||
| 5845 | " </tr>\n", |
||
| 5846 | " <tr>\n", |
||
| 5847 | " <th>371</th>\n", |
||
| 5848 | " <td>0</td>\n", |
||
| 5849 | " <td>0</td>\n", |
||
| 5850 | " <td>0</td>\n", |
||
| 5851 | " <td>0</td>\n", |
||
| 5852 | " <td>0</td>\n", |
||
| 5853 | " <td>0</td>\n", |
||
| 5854 | " <td>0</td>\n", |
||
| 5855 | " <td>0</td>\n", |
||
| 5856 | " <td>0</td>\n", |
||
| 5857 | " <td>0</td>\n", |
||
| 5858 | " <td>...</td>\n", |
||
| 5859 | " <td>-0.6853</td>\n", |
||
| 5860 | " <td>-1.0240</td>\n", |
||
| 5861 | " <td>-1.2890</td>\n", |
||
| 5862 | " <td>-1.5666</td>\n", |
||
| 5863 | " <td>-0.1270</td>\n", |
||
| 5864 | " <td>-1.4662</td>\n", |
||
| 5865 | " <td>0.3981</td>\n", |
||
| 5866 | " <td>-0.5976</td>\n", |
||
| 5867 | " <td>-1.3822</td>\n", |
||
| 5868 | " <td>-0.4157</td>\n", |
||
| 5869 | " </tr>\n", |
||
| 5870 | " <tr>\n", |
||
| 5871 | " <th>372</th>\n", |
||
| 5872 | " <td>0</td>\n", |
||
| 5873 | " <td>0</td>\n", |
||
| 5874 | " <td>0</td>\n", |
||
| 5875 | " <td>0</td>\n", |
||
| 5876 | " <td>0</td>\n", |
||
| 5877 | " <td>0</td>\n", |
||
| 5878 | " <td>0</td>\n", |
||
| 5879 | " <td>0</td>\n", |
||
| 5880 | " <td>0</td>\n", |
||
| 5881 | " <td>0</td>\n", |
||
| 5882 | " <td>...</td>\n", |
||
| 5883 | " <td>0.0517</td>\n", |
||
| 5884 | " <td>-0.3570</td>\n", |
||
| 5885 | " <td>-0.4843</td>\n", |
||
| 5886 | " <td>-0.3792</td>\n", |
||
| 5887 | " <td>-0.1964</td>\n", |
||
| 5888 | " <td>0.4200</td>\n", |
||
| 5889 | " <td>3.2547</td>\n", |
||
| 5890 | " <td>-0.1232</td>\n", |
||
| 5891 | " <td>3.4519</td>\n", |
||
| 5892 | " <td>-0.1962</td>\n", |
||
| 5893 | " </tr>\n", |
||
| 5894 | " </tbody>\n", |
||
| 5895 | "</table>\n", |
||
| 5896 | "<p>373 rows × 20395 columns</p>\n", |
||
| 5897 | "</div>" |
||
| 5898 | ], |
||
| 5899 | "text/plain": [ |
||
| 5900 | " NDUFS5_cnv MACF1_cnv RNA5SP44_cnv KIAA0754_cnv BMP8A_cnv PABPC4_cnv \\\n", |
||
| 5901 | "0 -1 -1 -1 -1 -1 -1 \n", |
||
| 5902 | "1 2 2 2 2 2 2 \n", |
||
| 5903 | "2 0 0 0 0 0 0 \n", |
||
| 5904 | "3 0 0 0 0 0 0 \n", |
||
| 5905 | "4 0 0 0 0 0 0 \n", |
||
| 5906 | ".. ... ... ... ... ... ... \n", |
||
| 5907 | "368 2 2 2 2 2 2 \n", |
||
| 5908 | "369 0 0 0 0 0 0 \n", |
||
| 5909 | "370 1 1 1 1 1 1 \n", |
||
| 5910 | "371 0 0 0 0 0 0 \n", |
||
| 5911 | "372 0 0 0 0 0 0 \n", |
||
| 5912 | "\n", |
||
| 5913 | " SNORA55_cnv HEYL_cnv HPCAL4_cnv NT5C1A_cnv ... ZWINT_rnaseq \\\n", |
||
| 5914 | "0 -1 -1 -1 -1 ... -0.8388 \n", |
||
| 5915 | "1 2 2 2 2 ... -0.1083 \n", |
||
| 5916 | "2 0 0 0 0 ... -0.4155 \n", |
||
| 5917 | "3 0 0 0 0 ... -0.8143 \n", |
||
| 5918 | "4 0 0 0 0 ... 0.0983 \n", |
||
| 5919 | ".. ... ... ... ... ... ... \n", |
||
| 5920 | "368 2 2 2 2 ... -0.0291 \n", |
||
| 5921 | "369 0 0 0 0 ... 0.0497 \n", |
||
| 5922 | "370 1 1 1 1 ... 0.3822 \n", |
||
| 5923 | "371 0 0 0 0 ... -0.6853 \n", |
||
| 5924 | "372 0 0 0 0 ... 0.0517 \n", |
||
| 5925 | "\n", |
||
| 5926 | " ZXDA_rnaseq ZXDB_rnaseq ZXDC_rnaseq ZYG11A_rnaseq ZYG11B_rnaseq \\\n", |
||
| 5927 | "0 4.1375 3.9664 1.8437 -0.3959 -0.2561 \n", |
||
| 5928 | "1 0.3393 0.2769 1.7320 -0.0975 2.6955 \n", |
||
| 5929 | "2 1.6846 0.7711 -0.3061 -0.5016 2.8548 \n", |
||
| 5930 | "3 0.8344 1.5075 3.6068 -0.5004 -0.0747 \n", |
||
| 5931 | "4 -0.7908 -0.0053 -0.0643 -0.3706 0.3870 \n", |
||
| 5932 | ".. ... ... ... ... ... \n", |
||
| 5933 | "368 -0.1058 -0.6721 0.2802 1.9504 -0.8784 \n", |
||
| 5934 | "369 0.3673 -0.2208 0.3034 3.2580 -0.2089 \n", |
||
| 5935 | "370 -0.7003 -0.7661 -1.7035 -0.5423 -0.3488 \n", |
||
| 5936 | "371 -1.0240 -1.2890 -1.5666 -0.1270 -1.4662 \n", |
||
| 5937 | "372 -0.3570 -0.4843 -0.3792 -0.1964 0.4200 \n", |
||
| 5938 | "\n", |
||
| 5939 | " ZYX_rnaseq ZZEF1_rnaseq ZZZ3_rnaseq TPTEP1_rnaseq \n", |
||
| 5940 | "0 -0.2866 1.8770 -0.3179 -0.3633 \n", |
||
| 5941 | "1 -0.6741 1.0323 1.2766 -0.3982 \n", |
||
| 5942 | "2 -0.6171 -0.8608 -0.0486 -0.3962 \n", |
||
| 5943 | "3 -0.2185 -0.4379 1.6913 1.7748 \n", |
||
| 5944 | "4 -0.5589 -0.5979 0.0047 -0.3548 \n", |
||
| 5945 | ".. ... ... ... ... \n", |
||
| 5946 | "368 0.9506 0.0607 1.1883 -0.3521 \n", |
||
| 5947 | "369 1.6053 -0.8746 -0.4491 -0.3450 \n", |
||
| 5948 | "370 1.3713 -0.4365 2.3456 -0.3866 \n", |
||
| 5949 | "371 0.3981 -0.5976 -1.3822 -0.4157 \n", |
||
| 5950 | "372 3.2547 -0.1232 3.4519 -0.1962 \n", |
||
| 5951 | "\n", |
||
| 5952 | "[373 rows x 20395 columns]" |
||
| 5953 | ] |
||
| 5954 | }, |
||
| 5955 | "execution_count": 68, |
||
| 5956 | "metadata": {}, |
||
| 5957 | "output_type": "execute_result" |
||
| 5958 | } |
||
| 5959 | ], |
||
| 5960 | "source": [ |
||
| 5961 | "genomic_features" |
||
| 5962 | ] |
||
| 5963 | }, |
||
| 5964 | { |
||
| 5965 | "cell_type": "code", |
||
| 5966 | "execution_count": 11, |
||
| 5967 | "metadata": {}, |
||
| 5968 | "outputs": [], |
||
| 5969 | "source": [ |
||
| 5970 | "if 'case_id' not in slide_data:\n", |
||
| 5971 | " slide_data.index = slide_data.index.str[:12]\n", |
||
| 5972 | " slide_data['case_id'] = slide_data.index\n", |
||
| 5973 | " slide_data = slide_data.reset_index(drop=True)" |
||
| 5974 | ] |
||
| 5975 | }, |
||
| 5976 | { |
||
| 5977 | "cell_type": "code", |
||
| 5978 | "execution_count": 14, |
||
| 5979 | "metadata": {}, |
||
| 5980 | "outputs": [], |
||
| 5981 | "source": [ |
||
| 5982 | "new_cols = list(slide_data.columns[-2:]) + list(slide_data.columns[:-2])\n", |
||
| 5983 | "slide_data = slide_data[new_cols]" |
||
| 5984 | ] |
||
| 5985 | }, |
||
| 5986 | { |
||
| 5987 | "cell_type": "code", |
||
| 5988 | "execution_count": 15, |
||
| 5989 | "metadata": {}, |
||
| 5990 | "outputs": [ |
||
| 5991 | { |
||
| 5992 | "data": { |
||
| 5993 | "text/html": [ |
||
| 5994 | "<div>\n", |
||
| 5995 | "<style scoped>\n", |
||
| 5996 | " .dataframe tbody tr th:only-of-type {\n", |
||
| 5997 | " vertical-align: middle;\n", |
||
| 5998 | " }\n", |
||
| 5999 | "\n", |
||
| 6000 | " .dataframe tbody tr th {\n", |
||
| 6001 | " vertical-align: top;\n", |
||
| 6002 | " }\n", |
||
| 6003 | "\n", |
||
| 6004 | " .dataframe thead th {\n", |
||
| 6005 | " text-align: right;\n", |
||
| 6006 | " }\n", |
||
| 6007 | "</style>\n", |
||
| 6008 | "<table border=\"1\" class=\"dataframe\">\n", |
||
| 6009 | " <thead>\n", |
||
| 6010 | " <tr style=\"text-align: right;\">\n", |
||
| 6011 | " <th></th>\n", |
||
| 6012 | " <th>ZZZ3_rnaseq</th>\n", |
||
| 6013 | " <th>TPTEP1_rnaseq</th>\n", |
||
| 6014 | " <th>slide_id</th>\n", |
||
| 6015 | " <th>site</th>\n", |
||
| 6016 | " <th>is_female</th>\n", |
||
| 6017 | " <th>oncotree_code</th>\n", |
||
| 6018 | " <th>age</th>\n", |
||
| 6019 | " <th>survival_months</th>\n", |
||
| 6020 | " <th>censorship</th>\n", |
||
| 6021 | " <th>train</th>\n", |
||
| 6022 | " <th>...</th>\n", |
||
| 6023 | " <th>ZW10_rnaseq</th>\n", |
||
| 6024 | " <th>ZWILCH_rnaseq</th>\n", |
||
| 6025 | " <th>ZWINT_rnaseq</th>\n", |
||
| 6026 | " <th>ZXDA_rnaseq</th>\n", |
||
| 6027 | " <th>ZXDB_rnaseq</th>\n", |
||
| 6028 | " <th>ZXDC_rnaseq</th>\n", |
||
| 6029 | " <th>ZYG11A_rnaseq</th>\n", |
||
| 6030 | " <th>ZYG11B_rnaseq</th>\n", |
||
| 6031 | " <th>ZYX_rnaseq</th>\n", |
||
| 6032 | " <th>ZZEF1_rnaseq</th>\n", |
||
| 6033 | " </tr>\n", |
||
| 6034 | " <tr>\n", |
||
| 6035 | " <th>case_id</th>\n", |
||
| 6036 | " <th></th>\n", |
||
| 6037 | " <th></th>\n", |
||
| 6038 | " <th></th>\n", |
||
| 6039 | " <th></th>\n", |
||
| 6040 | " <th></th>\n", |
||
| 6041 | " <th></th>\n", |
||
| 6042 | " <th></th>\n", |
||
| 6043 | " <th></th>\n", |
||
| 6044 | " <th></th>\n", |
||
| 6045 | " <th></th>\n", |
||
| 6046 | " <th></th>\n", |
||
| 6047 | " <th></th>\n", |
||
| 6048 | " <th></th>\n", |
||
| 6049 | " <th></th>\n", |
||
| 6050 | " <th></th>\n", |
||
| 6051 | " <th></th>\n", |
||
| 6052 | " <th></th>\n", |
||
| 6053 | " <th></th>\n", |
||
| 6054 | " <th></th>\n", |
||
| 6055 | " <th></th>\n", |
||
| 6056 | " <th></th>\n", |
||
| 6057 | " </tr>\n", |
||
| 6058 | " </thead>\n", |
||
| 6059 | " <tbody>\n", |
||
| 6060 | " <tr>\n", |
||
| 6061 | " <th>TCGA-2F-A9KO</th>\n", |
||
| 6062 | " <td>-0.3179</td>\n", |
||
| 6063 | " <td>-0.3633</td>\n", |
||
| 6064 | " <td>TCGA-2F-A9KO-01Z-00-DX1.195576CF-B739-4BD9-B15...</td>\n", |
||
| 6065 | " <td>2F</td>\n", |
||
| 6066 | " <td>0</td>\n", |
||
| 6067 | " <td>BLCA</td>\n", |
||
| 6068 | " <td>63</td>\n", |
||
| 6069 | " <td>24.11</td>\n", |
||
| 6070 | " <td>0</td>\n", |
||
| 6071 | " <td>1.0</td>\n", |
||
| 6072 | " <td>...</td>\n", |
||
| 6073 | " <td>-0.7172</td>\n", |
||
| 6074 | " <td>0.7409</td>\n", |
||
| 6075 | " <td>-0.8388</td>\n", |
||
| 6076 | " <td>4.1375</td>\n", |
||
| 6077 | " <td>3.9664</td>\n", |
||
| 6078 | " <td>1.8437</td>\n", |
||
| 6079 | " <td>-0.3959</td>\n", |
||
| 6080 | " <td>-0.2561</td>\n", |
||
| 6081 | " <td>-0.2866</td>\n", |
||
| 6082 | " <td>1.8770</td>\n", |
||
| 6083 | " </tr>\n", |
||
| 6084 | " <tr>\n", |
||
| 6085 | " <th>TCGA-2F-A9KP</th>\n", |
||
| 6086 | " <td>1.2766</td>\n", |
||
| 6087 | " <td>-0.3982</td>\n", |
||
| 6088 | " <td>TCGA-2F-A9KP-01Z-00-DX1.3CDF534E-958F-4467-AA7...</td>\n", |
||
| 6089 | " <td>2F</td>\n", |
||
| 6090 | " <td>0</td>\n", |
||
| 6091 | " <td>BLCA</td>\n", |
||
| 6092 | " <td>66</td>\n", |
||
| 6093 | " <td>11.96</td>\n", |
||
| 6094 | " <td>0</td>\n", |
||
| 6095 | " <td>1.0</td>\n", |
||
| 6096 | " <td>...</td>\n", |
||
| 6097 | " <td>0.6373</td>\n", |
||
| 6098 | " <td>0.8559</td>\n", |
||
| 6099 | " <td>-0.1083</td>\n", |
||
| 6100 | " <td>0.3393</td>\n", |
||
| 6101 | " <td>0.2769</td>\n", |
||
| 6102 | " <td>1.7320</td>\n", |
||
| 6103 | " <td>-0.0975</td>\n", |
||
| 6104 | " <td>2.6955</td>\n", |
||
| 6105 | " <td>-0.6741</td>\n", |
||
| 6106 | " <td>1.0323</td>\n", |
||
| 6107 | " </tr>\n", |
||
| 6108 | " <tr>\n", |
||
| 6109 | " <th>TCGA-2F-A9KP</th>\n", |
||
| 6110 | " <td>1.2766</td>\n", |
||
| 6111 | " <td>-0.3982</td>\n", |
||
| 6112 | " <td>TCGA-2F-A9KP-01Z-00-DX2.718C82A3-252B-498E-BFB...</td>\n", |
||
| 6113 | " <td>2F</td>\n", |
||
| 6114 | " <td>0</td>\n", |
||
| 6115 | " <td>BLCA</td>\n", |
||
| 6116 | " <td>66</td>\n", |
||
| 6117 | " <td>11.96</td>\n", |
||
| 6118 | " <td>0</td>\n", |
||
| 6119 | " <td>1.0</td>\n", |
||
| 6120 | " <td>...</td>\n", |
||
| 6121 | " <td>0.6373</td>\n", |
||
| 6122 | " <td>0.8559</td>\n", |
||
| 6123 | " <td>-0.1083</td>\n", |
||
| 6124 | " <td>0.3393</td>\n", |
||
| 6125 | " <td>0.2769</td>\n", |
||
| 6126 | " <td>1.7320</td>\n", |
||
| 6127 | " <td>-0.0975</td>\n", |
||
| 6128 | " <td>2.6955</td>\n", |
||
| 6129 | " <td>-0.6741</td>\n", |
||
| 6130 | " <td>1.0323</td>\n", |
||
| 6131 | " </tr>\n", |
||
| 6132 | " <tr>\n", |
||
| 6133 | " <th>TCGA-2F-A9KQ</th>\n", |
||
| 6134 | " <td>-0.0486</td>\n", |
||
| 6135 | " <td>-0.3962</td>\n", |
||
| 6136 | " <td>TCGA-2F-A9KQ-01Z-00-DX1.1C8CB2DD-5CC6-4E99-A0F...</td>\n", |
||
| 6137 | " <td>2F</td>\n", |
||
| 6138 | " <td>0</td>\n", |
||
| 6139 | " <td>BLCA</td>\n", |
||
| 6140 | " <td>69</td>\n", |
||
| 6141 | " <td>94.81</td>\n", |
||
| 6142 | " <td>1</td>\n", |
||
| 6143 | " <td>1.0</td>\n", |
||
| 6144 | " <td>...</td>\n", |
||
| 6145 | " <td>-0.5676</td>\n", |
||
| 6146 | " <td>-0.0621</td>\n", |
||
| 6147 | " <td>-0.4155</td>\n", |
||
| 6148 | " <td>1.6846</td>\n", |
||
| 6149 | " <td>0.7711</td>\n", |
||
| 6150 | " <td>-0.3061</td>\n", |
||
| 6151 | " <td>-0.5016</td>\n", |
||
| 6152 | " <td>2.8548</td>\n", |
||
| 6153 | " <td>-0.6171</td>\n", |
||
| 6154 | " <td>-0.8608</td>\n", |
||
| 6155 | " </tr>\n", |
||
| 6156 | " <tr>\n", |
||
| 6157 | " <th>TCGA-2F-A9KR</th>\n", |
||
| 6158 | " <td>1.6913</td>\n", |
||
| 6159 | " <td>1.7748</td>\n", |
||
| 6160 | " <td>TCGA-2F-A9KR-01Z-00-DX1.D6A4BD2D-18F3-4FA6-827...</td>\n", |
||
| 6161 | " <td>2F</td>\n", |
||
| 6162 | " <td>1</td>\n", |
||
| 6163 | " <td>BLCA</td>\n", |
||
| 6164 | " <td>59</td>\n", |
||
| 6165 | " <td>104.57</td>\n", |
||
| 6166 | " <td>0</td>\n", |
||
| 6167 | " <td>1.0</td>\n", |
||
| 6168 | " <td>...</td>\n", |
||
| 6169 | " <td>-1.3825</td>\n", |
||
| 6170 | " <td>0.3550</td>\n", |
||
| 6171 | " <td>-0.8143</td>\n", |
||
| 6172 | " <td>0.8344</td>\n", |
||
| 6173 | " <td>1.5075</td>\n", |
||
| 6174 | " <td>3.6068</td>\n", |
||
| 6175 | " <td>-0.5004</td>\n", |
||
| 6176 | " <td>-0.0747</td>\n", |
||
| 6177 | " <td>-0.2185</td>\n", |
||
| 6178 | " <td>-0.4379</td>\n", |
||
| 6179 | " </tr>\n", |
||
| 6180 | " <tr>\n", |
||
| 6181 | " <th>...</th>\n", |
||
| 6182 | " <td>...</td>\n", |
||
| 6183 | " <td>...</td>\n", |
||
| 6184 | " <td>...</td>\n", |
||
| 6185 | " <td>...</td>\n", |
||
| 6186 | " <td>...</td>\n", |
||
| 6187 | " <td>...</td>\n", |
||
| 6188 | " <td>...</td>\n", |
||
| 6189 | " <td>...</td>\n", |
||
| 6190 | " <td>...</td>\n", |
||
| 6191 | " <td>...</td>\n", |
||
| 6192 | " <td>...</td>\n", |
||
| 6193 | " <td>...</td>\n", |
||
| 6194 | " <td>...</td>\n", |
||
| 6195 | " <td>...</td>\n", |
||
| 6196 | " <td>...</td>\n", |
||
| 6197 | " <td>...</td>\n", |
||
| 6198 | " <td>...</td>\n", |
||
| 6199 | " <td>...</td>\n", |
||
| 6200 | " <td>...</td>\n", |
||
| 6201 | " <td>...</td>\n", |
||
| 6202 | " <td>...</td>\n", |
||
| 6203 | " </tr>\n", |
||
| 6204 | " <tr>\n", |
||
| 6205 | " <th>TCGA-ZF-AA54</th>\n", |
||
| 6206 | " <td>1.1883</td>\n", |
||
| 6207 | " <td>-0.3521</td>\n", |
||
| 6208 | " <td>TCGA-ZF-AA54-01Z-00-DX1.9118BB51-333A-4257-A79...</td>\n", |
||
| 6209 | " <td>ZF</td>\n", |
||
| 6210 | " <td>0</td>\n", |
||
| 6211 | " <td>BLCA</td>\n", |
||
| 6212 | " <td>71</td>\n", |
||
| 6213 | " <td>19.38</td>\n", |
||
| 6214 | " <td>0</td>\n", |
||
| 6215 | " <td>1.0</td>\n", |
||
| 6216 | " <td>...</td>\n", |
||
| 6217 | " <td>-0.0898</td>\n", |
||
| 6218 | " <td>2.1092</td>\n", |
||
| 6219 | " <td>-0.0291</td>\n", |
||
| 6220 | " <td>-0.1058</td>\n", |
||
| 6221 | " <td>-0.6721</td>\n", |
||
| 6222 | " <td>0.2802</td>\n", |
||
| 6223 | " <td>1.9504</td>\n", |
||
| 6224 | " <td>-0.8784</td>\n", |
||
| 6225 | " <td>0.9506</td>\n", |
||
| 6226 | " <td>0.0607</td>\n", |
||
| 6227 | " </tr>\n", |
||
| 6228 | " <tr>\n", |
||
| 6229 | " <th>TCGA-ZF-AA58</th>\n", |
||
| 6230 | " <td>-0.4491</td>\n", |
||
| 6231 | " <td>-0.3450</td>\n", |
||
| 6232 | " <td>TCGA-ZF-AA58-01Z-00-DX1.85C3611E-11FA-4AAE-B88...</td>\n", |
||
| 6233 | " <td>ZF</td>\n", |
||
| 6234 | " <td>1</td>\n", |
||
| 6235 | " <td>BLCA</td>\n", |
||
| 6236 | " <td>61</td>\n", |
||
| 6237 | " <td>54.17</td>\n", |
||
| 6238 | " <td>1</td>\n", |
||
| 6239 | " <td>1.0</td>\n", |
||
| 6240 | " <td>...</td>\n", |
||
| 6241 | " <td>-0.2075</td>\n", |
||
| 6242 | " <td>-0.0617</td>\n", |
||
| 6243 | " <td>0.0497</td>\n", |
||
| 6244 | " <td>0.3673</td>\n", |
||
| 6245 | " <td>-0.2208</td>\n", |
||
| 6246 | " <td>0.3034</td>\n", |
||
| 6247 | " <td>3.2580</td>\n", |
||
| 6248 | " <td>-0.2089</td>\n", |
||
| 6249 | " <td>1.6053</td>\n", |
||
| 6250 | " <td>-0.8746</td>\n", |
||
| 6251 | " </tr>\n", |
||
| 6252 | " <tr>\n", |
||
| 6253 | " <th>TCGA-ZF-AA5H</th>\n", |
||
| 6254 | " <td>2.3456</td>\n", |
||
| 6255 | " <td>-0.3866</td>\n", |
||
| 6256 | " <td>TCGA-ZF-AA5H-01Z-00-DX1.2B5DF00E-E0FD-4C58-A82...</td>\n", |
||
| 6257 | " <td>ZF</td>\n", |
||
| 6258 | " <td>1</td>\n", |
||
| 6259 | " <td>BLCA</td>\n", |
||
| 6260 | " <td>60</td>\n", |
||
| 6261 | " <td>29.47</td>\n", |
||
| 6262 | " <td>1</td>\n", |
||
| 6263 | " <td>1.0</td>\n", |
||
| 6264 | " <td>...</td>\n", |
||
| 6265 | " <td>-1.4118</td>\n", |
||
| 6266 | " <td>-0.1236</td>\n", |
||
| 6267 | " <td>0.3822</td>\n", |
||
| 6268 | " <td>-0.7003</td>\n", |
||
| 6269 | " <td>-0.7661</td>\n", |
||
| 6270 | " <td>-1.7035</td>\n", |
||
| 6271 | " <td>-0.5423</td>\n", |
||
| 6272 | " <td>-0.3488</td>\n", |
||
| 6273 | " <td>1.3713</td>\n", |
||
| 6274 | " <td>-0.4365</td>\n", |
||
| 6275 | " </tr>\n", |
||
| 6276 | " <tr>\n", |
||
| 6277 | " <th>TCGA-ZF-AA5N</th>\n", |
||
| 6278 | " <td>-1.3822</td>\n", |
||
| 6279 | " <td>-0.4157</td>\n", |
||
| 6280 | " <td>TCGA-ZF-AA5N-01Z-00-DX1.A207E3EE-CC7D-4267-A77...</td>\n", |
||
| 6281 | " <td>ZF</td>\n", |
||
| 6282 | " <td>1</td>\n", |
||
| 6283 | " <td>BLCA</td>\n", |
||
| 6284 | " <td>62</td>\n", |
||
| 6285 | " <td>5.52</td>\n", |
||
| 6286 | " <td>0</td>\n", |
||
| 6287 | " <td>1.0</td>\n", |
||
| 6288 | " <td>...</td>\n", |
||
| 6289 | " <td>-0.1733</td>\n", |
||
| 6290 | " <td>-0.2397</td>\n", |
||
| 6291 | " <td>-0.6853</td>\n", |
||
| 6292 | " <td>-1.0240</td>\n", |
||
| 6293 | " <td>-1.2890</td>\n", |
||
| 6294 | " <td>-1.5666</td>\n", |
||
| 6295 | " <td>-0.1270</td>\n", |
||
| 6296 | " <td>-1.4662</td>\n", |
||
| 6297 | " <td>0.3981</td>\n", |
||
| 6298 | " <td>-0.5976</td>\n", |
||
| 6299 | " </tr>\n", |
||
| 6300 | " <tr>\n", |
||
| 6301 | " <th>TCGA-ZF-AA5P</th>\n", |
||
| 6302 | " <td>3.4519</td>\n", |
||
| 6303 | " <td>-0.1962</td>\n", |
||
| 6304 | " <td>TCGA-ZF-AA5P-01Z-00-DX1.B91697A2-A186-4E67-A81...</td>\n", |
||
| 6305 | " <td>ZF</td>\n", |
||
| 6306 | " <td>0</td>\n", |
||
| 6307 | " <td>BLCA</td>\n", |
||
| 6308 | " <td>65</td>\n", |
||
| 6309 | " <td>12.22</td>\n", |
||
| 6310 | " <td>1</td>\n", |
||
| 6311 | " <td>1.0</td>\n", |
||
| 6312 | " <td>...</td>\n", |
||
| 6313 | " <td>-1.1056</td>\n", |
||
| 6314 | " <td>-0.6634</td>\n", |
||
| 6315 | " <td>0.0517</td>\n", |
||
| 6316 | " <td>-0.3570</td>\n", |
||
| 6317 | " <td>-0.4843</td>\n", |
||
| 6318 | " <td>-0.3792</td>\n", |
||
| 6319 | " <td>-0.1964</td>\n", |
||
| 6320 | " <td>0.4200</td>\n", |
||
| 6321 | " <td>3.2547</td>\n", |
||
| 6322 | " <td>-0.1232</td>\n", |
||
| 6323 | " </tr>\n", |
||
| 6324 | " </tbody>\n", |
||
| 6325 | "</table>\n", |
||
| 6326 | "<p>437 rows × 20403 columns</p>\n", |
||
| 6327 | "</div>" |
||
| 6328 | ], |
||
| 6329 | "text/plain": [ |
||
| 6330 | " ZZZ3_rnaseq TPTEP1_rnaseq \\\n", |
||
| 6331 | "case_id \n", |
||
| 6332 | "TCGA-2F-A9KO -0.3179 -0.3633 \n", |
||
| 6333 | "TCGA-2F-A9KP 1.2766 -0.3982 \n", |
||
| 6334 | "TCGA-2F-A9KP 1.2766 -0.3982 \n", |
||
| 6335 | "TCGA-2F-A9KQ -0.0486 -0.3962 \n", |
||
| 6336 | "TCGA-2F-A9KR 1.6913 1.7748 \n", |
||
| 6337 | "... ... ... \n", |
||
| 6338 | "TCGA-ZF-AA54 1.1883 -0.3521 \n", |
||
| 6339 | "TCGA-ZF-AA58 -0.4491 -0.3450 \n", |
||
| 6340 | "TCGA-ZF-AA5H 2.3456 -0.3866 \n", |
||
| 6341 | "TCGA-ZF-AA5N -1.3822 -0.4157 \n", |
||
| 6342 | "TCGA-ZF-AA5P 3.4519 -0.1962 \n", |
||
| 6343 | "\n", |
||
| 6344 | " slide_id site \\\n", |
||
| 6345 | "case_id \n", |
||
| 6346 | "TCGA-2F-A9KO TCGA-2F-A9KO-01Z-00-DX1.195576CF-B739-4BD9-B15... 2F \n", |
||
| 6347 | "TCGA-2F-A9KP TCGA-2F-A9KP-01Z-00-DX1.3CDF534E-958F-4467-AA7... 2F \n", |
||
| 6348 | "TCGA-2F-A9KP TCGA-2F-A9KP-01Z-00-DX2.718C82A3-252B-498E-BFB... 2F \n", |
||
| 6349 | "TCGA-2F-A9KQ TCGA-2F-A9KQ-01Z-00-DX1.1C8CB2DD-5CC6-4E99-A0F... 2F \n", |
||
| 6350 | "TCGA-2F-A9KR TCGA-2F-A9KR-01Z-00-DX1.D6A4BD2D-18F3-4FA6-827... 2F \n", |
||
| 6351 | "... ... ... \n", |
||
| 6352 | "TCGA-ZF-AA54 TCGA-ZF-AA54-01Z-00-DX1.9118BB51-333A-4257-A79... ZF \n", |
||
| 6353 | "TCGA-ZF-AA58 TCGA-ZF-AA58-01Z-00-DX1.85C3611E-11FA-4AAE-B88... ZF \n", |
||
| 6354 | "TCGA-ZF-AA5H TCGA-ZF-AA5H-01Z-00-DX1.2B5DF00E-E0FD-4C58-A82... ZF \n", |
||
| 6355 | "TCGA-ZF-AA5N TCGA-ZF-AA5N-01Z-00-DX1.A207E3EE-CC7D-4267-A77... ZF \n", |
||
| 6356 | "TCGA-ZF-AA5P TCGA-ZF-AA5P-01Z-00-DX1.B91697A2-A186-4E67-A81... ZF \n", |
||
| 6357 | "\n", |
||
| 6358 | " is_female oncotree_code age survival_months censorship \\\n", |
||
| 6359 | "case_id \n", |
||
| 6360 | "TCGA-2F-A9KO 0 BLCA 63 24.11 0 \n", |
||
| 6361 | "TCGA-2F-A9KP 0 BLCA 66 11.96 0 \n", |
||
| 6362 | "TCGA-2F-A9KP 0 BLCA 66 11.96 0 \n", |
||
| 6363 | "TCGA-2F-A9KQ 0 BLCA 69 94.81 1 \n", |
||
| 6364 | "TCGA-2F-A9KR 1 BLCA 59 104.57 0 \n", |
||
| 6365 | "... ... ... ... ... ... \n", |
||
| 6366 | "TCGA-ZF-AA54 0 BLCA 71 19.38 0 \n", |
||
| 6367 | "TCGA-ZF-AA58 1 BLCA 61 54.17 1 \n", |
||
| 6368 | "TCGA-ZF-AA5H 1 BLCA 60 29.47 1 \n", |
||
| 6369 | "TCGA-ZF-AA5N 1 BLCA 62 5.52 0 \n", |
||
| 6370 | "TCGA-ZF-AA5P 0 BLCA 65 12.22 1 \n", |
||
| 6371 | "\n", |
||
| 6372 | " train ... ZW10_rnaseq ZWILCH_rnaseq ZWINT_rnaseq \\\n", |
||
| 6373 | "case_id ... \n", |
||
| 6374 | "TCGA-2F-A9KO 1.0 ... -0.7172 0.7409 -0.8388 \n", |
||
| 6375 | "TCGA-2F-A9KP 1.0 ... 0.6373 0.8559 -0.1083 \n", |
||
| 6376 | "TCGA-2F-A9KP 1.0 ... 0.6373 0.8559 -0.1083 \n", |
||
| 6377 | "TCGA-2F-A9KQ 1.0 ... -0.5676 -0.0621 -0.4155 \n", |
||
| 6378 | "TCGA-2F-A9KR 1.0 ... -1.3825 0.3550 -0.8143 \n", |
||
| 6379 | "... ... ... ... ... ... \n", |
||
| 6380 | "TCGA-ZF-AA54 1.0 ... -0.0898 2.1092 -0.0291 \n", |
||
| 6381 | "TCGA-ZF-AA58 1.0 ... -0.2075 -0.0617 0.0497 \n", |
||
| 6382 | "TCGA-ZF-AA5H 1.0 ... -1.4118 -0.1236 0.3822 \n", |
||
| 6383 | "TCGA-ZF-AA5N 1.0 ... -0.1733 -0.2397 -0.6853 \n", |
||
| 6384 | "TCGA-ZF-AA5P 1.0 ... -1.1056 -0.6634 0.0517 \n", |
||
| 6385 | "\n", |
||
| 6386 | " ZXDA_rnaseq ZXDB_rnaseq ZXDC_rnaseq ZYG11A_rnaseq \\\n", |
||
| 6387 | "case_id \n", |
||
| 6388 | "TCGA-2F-A9KO 4.1375 3.9664 1.8437 -0.3959 \n", |
||
| 6389 | "TCGA-2F-A9KP 0.3393 0.2769 1.7320 -0.0975 \n", |
||
| 6390 | "TCGA-2F-A9KP 0.3393 0.2769 1.7320 -0.0975 \n", |
||
| 6391 | "TCGA-2F-A9KQ 1.6846 0.7711 -0.3061 -0.5016 \n", |
||
| 6392 | "TCGA-2F-A9KR 0.8344 1.5075 3.6068 -0.5004 \n", |
||
| 6393 | "... ... ... ... ... \n", |
||
| 6394 | "TCGA-ZF-AA54 -0.1058 -0.6721 0.2802 1.9504 \n", |
||
| 6395 | "TCGA-ZF-AA58 0.3673 -0.2208 0.3034 3.2580 \n", |
||
| 6396 | "TCGA-ZF-AA5H -0.7003 -0.7661 -1.7035 -0.5423 \n", |
||
| 6397 | "TCGA-ZF-AA5N -1.0240 -1.2890 -1.5666 -0.1270 \n", |
||
| 6398 | "TCGA-ZF-AA5P -0.3570 -0.4843 -0.3792 -0.1964 \n", |
||
| 6399 | "\n", |
||
| 6400 | " ZYG11B_rnaseq ZYX_rnaseq ZZEF1_rnaseq \n", |
||
| 6401 | "case_id \n", |
||
| 6402 | "TCGA-2F-A9KO -0.2561 -0.2866 1.8770 \n", |
||
| 6403 | "TCGA-2F-A9KP 2.6955 -0.6741 1.0323 \n", |
||
| 6404 | "TCGA-2F-A9KP 2.6955 -0.6741 1.0323 \n", |
||
| 6405 | "TCGA-2F-A9KQ 2.8548 -0.6171 -0.8608 \n", |
||
| 6406 | "TCGA-2F-A9KR -0.0747 -0.2185 -0.4379 \n", |
||
| 6407 | "... ... ... ... \n", |
||
| 6408 | "TCGA-ZF-AA54 -0.8784 0.9506 0.0607 \n", |
||
| 6409 | "TCGA-ZF-AA58 -0.2089 1.6053 -0.8746 \n", |
||
| 6410 | "TCGA-ZF-AA5H -0.3488 1.3713 -0.4365 \n", |
||
| 6411 | "TCGA-ZF-AA5N -1.4662 0.3981 -0.5976 \n", |
||
| 6412 | "TCGA-ZF-AA5P 0.4200 3.2547 -0.1232 \n", |
||
| 6413 | "\n", |
||
| 6414 | "[437 rows x 20403 columns]" |
||
| 6415 | ] |
||
| 6416 | }, |
||
| 6417 | "execution_count": 15, |
||
| 6418 | "metadata": {}, |
||
| 6419 | "output_type": "execute_result" |
||
| 6420 | } |
||
| 6421 | ], |
||
| 6422 | "source": [ |
||
| 6423 | "slide_data" |
||
| 6424 | ] |
||
| 6425 | } |
||
| 6426 | ], |
||
| 6427 | "metadata": { |
||
| 6428 | "kernelspec": { |
||
| 6429 | "display_name": "Python 3", |
||
| 6430 | "language": "python", |
||
| 6431 | "name": "python3" |
||
| 6432 | }, |
||
| 6433 | "language_info": { |
||
| 6434 | "codemirror_mode": { |
||
| 6435 | "name": "ipython", |
||
| 6436 | "version": 3 |
||
| 6437 | }, |
||
| 6438 | "file_extension": ".py", |
||
| 6439 | "mimetype": "text/x-python", |
||
| 6440 | "name": "python", |
||
| 6441 | "nbconvert_exporter": "python", |
||
| 6442 | "pygments_lexer": "ipython3", |
||
| 6443 | "version": "3.7.7" |
||
| 6444 | } |
||
| 6445 | }, |
||
| 6446 | "nbformat": 4, |
||
| 6447 | "nbformat_minor": 4 |
||
| 6448 | } |