Switch to unified view

a/README.md b/README.md
1
# PathCNN: Interpretable convolutional neural networks for survival prediction and pathway analysis applied to glioblastoma
1
# PathCNN: Interpretable convolutional neural networks for survival prediction and pathway analysis applied to glioblastoma
2
2
3
Jung Hun Oh <sup>1,†,∗</sup>, Wookjin Choi <sup>2,†</sup>, Euiseong Ko <sup>3</sup>, Mingon Kang <sup>3,∗</sup>, Allen Tannenbaum <sup>4</sup> and Joseph O. Deasy <sup>1</sup>  
3
Jung Hun Oh <sup>1,†,∗</sup>, Wookjin Choi <sup>2,†</sup>, Euiseong Ko <sup>3</sup>, Mingon Kang <sup>3,∗</sup>, Allen Tannenbaum <sup>4</sup> and Joseph O. Deasy <sup>1</sup>  
4
4
5
<sup>1</sup>Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, USA, 
5
<sup>1</sup>Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, USA, 
6
<sup>2</sup>Department of Computer Science, Virginia State University, Petersburg, USA, 
6
<sup>2</sup>Department of Computer Science, Virginia State University, Petersburg, USA, 
7
<sup>3</sup>Department of Computer Science, University of Nevada, Las Vegas, USA and 
7
<sup>3</sup>Department of Computer Science, University of Nevada, Las Vegas, USA and 
8
<sup>4</sup>Departments of Computer Science and Applied Mathematics & Statistics, Stony Brook University, New York, USA  
8
<sup>4</sup>Departments of Computer Science and Applied Mathematics & Statistics, Stony Brook University, New York, USA  
9
9
10
<sup>*</sup>To whom correspondence should be addressed.  
10
<sup>*</sup>To whom correspondence should be addressed.  
11
<sup>†</sup>The authors wish it to be known that, in their opinion, the first two authors should be regarded as Joint First Authors.  
11
<sup>†</sup>The authors wish it to be known that, in their opinion, the first two authors should be regarded as Joint First Authors.  
12
12
13
Contact: <ohj@mskcc.org> or <mingon.kang@unlv.edu>
13
Contact: <ohj@mskcc.org> or <mingon.kang@unlv.edu>
14
14
15
![PathCNN](img/pathcnn.png)
16
17
1. Model Building  
15
1. Model Building  
18
   - PathCNN.py  
16
   - PathCNN.py  
19
17
20
2. GradCAM  
18
2. GradCAM  
21
   - PathCNN_GradCAM_modeling.py: to generate a model for GradCAM (PathCNN_model.h5)
19
   - PathCNN_GradCAM_modeling.py: to generate a model for GradCAM (PathCNN_model.h5)
22
   - PathCNN_GradCAM.py: to generate GradCAM images and a resultant file (pathcnn_gradcam.csv)
20
   - PathCNN_GradCAM.py: to generate GradCAM images and a resultant file (pathcnn_gradcam.csv)
23
21
24
3. Multi-omics data
22
3. Multi-omics data
25
   - GBM multi-omics data including mRNA expression, CNV, and DNA methylation were downloaded from the CBioPortal database.
23
   - GBM multi-omics data including mRNA expression, CNV, and DNA methylation were downloaded from the CBioPortal database.
26
   - Pathway information was downloaded from the KEGG database.
24
   - Pathway information was downloaded from the KEGG database.
27
   - PCA was performed for each pathway in individual omics types.
25
   - PCA was performed for each pathway in individual omics types.
28
   
26
   
29
   Five PCs in each omics type are in the following files:
27
   Five PCs in each omics type are in the following files:
30
   - PCA_EXP.xlsx, PCA_CNV.xlsx, PCA_MT.xlsx
28
   - PCA_EXP.xlsx, PCA_CNV.xlsx, PCA_MT.xlsx
31
   
29
   
32
   Clinival variables are in the following file:
30
   Clinival variables are in the following file:
33
   - Clinical.xlsx
31
   - Clinical.xlsx