AItrika / Git / Diff of /README.md

Models:

philipB/

AItrika

Downloads: 1

Diff of /README.md [1bdb11] .. [ed1001]

Switch to unified view


# AItrika



[![License](https://img.shields.io/badge/License-Apache%202.0-orange.svg)](https://opensource.org/licenses/Apache-2.0)
![GitHub forks](https://img.shields.io/github/forks/dSupertramp/AItrika)
![GitHub commit activity (branch)](https://img.shields.io/github/commit-activity/t/dSupertramp/AItrika/main)
![GitHub last commit (branch)](https://img.shields.io/github/last-commit/dSupertramp/AItrika/main)

![Static Badge](https://img.shields.io/badge/medical-content?logo=syringe&logoColor=cyan&color=cyan)

Enhance your knowledge in medical research.

AItrika (formerly **PubGPT**) is a tool that can extract lots of relevant informations inside medical papers in an easy way:

- Abstract
- Full text (when available)
- Genes
- Diseases
- Mutations
- Associations between genes and diseases
- MeSH terms
- Other terms
- Results
- Bibliography

And so on!

## 🚀 Run the demo app

You can try AItrika with the Streamlit app by running:

```
streamlit run app.py
```

Or you can use it a script by running:

```
python main.py
```

## 📦 Install

To install everything, you need `uv`.

First of all, install `uv` with the command:

```
python main.py
```

After that, create a virtual environment with the command:

```
uv venv venv_name
```

Activate the virtual env:

```
source venv_name/bin/activate
```

And install dependencies:

```
uv pip install -r requirements.in
```

## 🔑 Set LLM API Keys

In order to set API keys, insert your keys into the `env.example` file and rename it to `.env`.

## 🔍 Usage

You can easily get informations of a paper by passing a PubMed ID:

```python
from aitrika.engine.aitrika import OnlineAItrika
aitrika_engine = OnlineAItrika(pubmed_id=pubmed_id)
title = aitrika_engine.get_title()
print(title)
```

Or you can parse a local pdf:

```python
from aitrika.engine.aitrika import LocalAItrika
aitrika_engine = LocalAItrika(pdf_path = pdf_path)
title = aitrika_engine.get_title()
print(title)
```

```
Breast cancer genes: beyond BRCA1 and BRCA2.
```

You can get other informations, like the associations between genes and diseases:

```python
associations = aitrika_engine.get_associations()
```

```
[
  {
    "gene": "BRIP1",
    "disease": "Breast Neoplasms"
  },
  {
    "gene": "PTEN",
    "disease": "Breast Neoplasms"
  },
  {
    "gene": "CHEK2",
    "disease": "Breast Neoplasms"
  },
]
...
```

Or you can get a nice formatted DataFrame:

```python
associations = aitrika_engine.associations(dataframe = True)
```

```
      gene                          disease
0    BRIP1                 Breast Neoplasms
1     PTEN                 Breast Neoplasms
2    CHEK2                 Breast Neoplasms
...
```

With the power of RAG, you can query your document:

```python
## Prepare the documents
documents = generate_documents(content=abstract)

## Set the LLM
llm = GroqLLM(documents=documents, api_key=os.getenv("GROQ_API_KEY"))

## Query your document
query = "Is BRCA1 associated with breast cancer?"
print(llm.query(query=query))
```

```
The provided text suggests that BRCA1 is associated with breast cancer, as it is listed among the high-penetrance genes identified in family linkage studies as responsible for inherited syndromes of breast cancer.
```

Or you can extract other informations:

```python
results = engine.extract_results(llm=llm)
print(results)
```

```
** RESULTS **

- High-penetrance genes - BRCA1, BRCA2, PTEN, TP53 - responsible for inherited syndromes
- Moderate-penetrance genes - CHEK2, ATM, BRIP1, PALB2, RAD51C - associated with moderate BC risk
- Low-penetrance alleles - common alleles - associated with slightly increased or decreased risk of BC
- Current clinical practice - high-penetrance genes - widely used
- Future prospect - all familial breast cancer genes - to be included in genetic test
- Research need - clinical management - of moderate and low-risk variants
```

## 🚀 Run the API

To run the AItrika API, follow these steps:

1. Ensure you have set up your environment and installed all dependencies as described in the Installation section.

2. Run the API server using the following command:

```bash
python api.py
```

The API will start running on http://0.0.0.0:8000. You can now make requests to the various endpoints:

- /associations: Get associations from a PubMed article
- /abstract: Get abstract of a PubMed article
- /query: Query a PubMed article
- /results: Get results from a PubMed article
- /participants: Get number of participants from a PubMed article
- /outcomes: Get outcomes from a PubMed article

You can use tools like curl, Postman, or any HTTP client to interact with the API. For example:

```bash
curl -X POST "http://localhost:8000/abstract" -H "Content-Type: application/json" -d '{"pubmed_id": 12345678}'
```

The API documentation is automatically generated and saved to <code>docs/api-reference/openapi.json</code>.
You can use this file with tools like Swagger UI for a more interactive API exploration experience.

## Support the Project

If you find this project useful, please consider supporting it:

- 🌟 Star the project on GitHub
- 🐛 Report bugs or suggest new features
- 🤝 Contribute with pull requests
- ☕️ [Buy me a coffee](https://www.buymeacoffee.com/dsupertramp) or consider a sponsor.

### Commercial / Business use

If you're using this project in a business or commercial context, please [contact me](salvatoredanilopalumbo@gmail.com).

I'm available for consulting, custom development, or commercial licensing.

Your support helps keep this project active and continuously improving. Thank you!

## License

AItrika is licensed under the Apache 2.0 License. See the LICENSE file for more details.

## Star History

[![Star History Chart](https://api.star-history.com/svg?repos=dSupertramp/AItrika&type=Date)](https://star-history.com/#dSupertramp/AItrika&Date)

	a/README.md		b/README.md
1	# AItrika	1	# AItrika
2		2
3	![AItrika](images/logo.png)
4
5	[![License](https://img.shields.io/badge/License-Apache%202.0-orange.svg)](https://opensource.org/licenses/Apache-2.0)	3	[![License](https://img.shields.io/badge/License-Apache%202.0-orange.svg)](https://opensource.org/licenses/Apache-2.0)
6	![GitHub forks](https://img.shields.io/github/forks/dSupertramp/AItrika)	4	![GitHub forks](https://img.shields.io/github/forks/dSupertramp/AItrika)
7	![GitHub commit activity (branch)](https://img.shields.io/github/commit-activity/t/dSupertramp/AItrika/main)	5	![GitHub commit activity (branch)](https://img.shields.io/github/commit-activity/t/dSupertramp/AItrika/main)
8	![GitHub last commit (branch)](https://img.shields.io/github/last-commit/dSupertramp/AItrika/main)	6	![GitHub last commit (branch)](https://img.shields.io/github/last-commit/dSupertramp/AItrika/main)
9		7
10	![Static Badge](https://img.shields.io/badge/medical-content?logo=syringe&logoColor=cyan&color=cyan)	8	![Static Badge](https://img.shields.io/badge/medical-content?logo=syringe&logoColor=cyan&color=cyan)
11		9
12	Enhance your knowledge in medical research.	10	Enhance your knowledge in medical research.
13		11
14	AItrika (formerly PubGPT) is a tool that can extract lots of relevant informations inside medical papers in an easy way:	12	AItrika (formerly PubGPT) is a tool that can extract lots of relevant informations inside medical papers in an easy way:
15		13
16	- Abstract	14	- Abstract
17	- Full text (when available)	15	- Full text (when available)
18	- Genes	16	- Genes
19	- Diseases	17	- Diseases
20	- Mutations	18	- Mutations
21	- Associations between genes and diseases	19	- Associations between genes and diseases
22	- MeSH terms	20	- MeSH terms
23	- Other terms	21	- Other terms
24	- Results	22	- Results
25	- Bibliography	23	- Bibliography
26		24
27	And so on!	25	And so on!
28		26
29	## 🚀 Run the demo app	27	## 🚀 Run the demo app
30		28
31	You can try AItrika with the Streamlit app by running:	29	You can try AItrika with the Streamlit app by running:
32		30
33	```	31	```
34	streamlit run app.py	32	streamlit run app.py
35	```	33	```
36		34
37	Or you can use it a script by running:	35	Or you can use it a script by running:
38		36
39	```	37	```
40	python main.py	38	python main.py
41	```	39	```
42		40
43	## 📦 Install	41	## 📦 Install
44		42
45	To install everything, you need `uv`.	43	To install everything, you need `uv`.
46		44
47	First of all, install `uv` with the command:	45	First of all, install `uv` with the command:
48		46
49	```	47	```
50	python main.py	48	python main.py
51	```	49	```
52		50
53	After that, create a virtual environment with the command:	51	After that, create a virtual environment with the command:
54		52
55	```	53	```
56	uv venv venv_name	54	uv venv venv_name
57	```	55	```
58		56
59	Activate the virtual env:	57	Activate the virtual env:
60		58
61	```	59	```
62	source venv_name/bin/activate	60	source venv_name/bin/activate
63	```	61	```
64		62
65	And install dependencies:	63	And install dependencies:
66		64
67	```	65	```
68	uv pip install -r requirements.in	66	uv pip install -r requirements.in
69	```	67	```
70		68
71	## 🔑 Set LLM API Keys	69	## 🔑 Set LLM API Keys
72		70
73	In order to set API keys, insert your keys into the `env.example` file and rename it to `.env`.	71	In order to set API keys, insert your keys into the `env.example` file and rename it to `.env`.
74		72
75	## 🔍 Usage	73	## 🔍 Usage
76		74
77	You can easily get informations of a paper by passing a PubMed ID:	75	You can easily get informations of a paper by passing a PubMed ID:
78		76
79	```python	77	```python
80	from aitrika.engine.aitrika import OnlineAItrika	78	from aitrika.engine.aitrika import OnlineAItrika
81	aitrika_engine = OnlineAItrika(pubmed_id=pubmed_id)	79	aitrika_engine = OnlineAItrika(pubmed_id=pubmed_id)
82	title = aitrika_engine.get_title()	80	title = aitrika_engine.get_title()
83	print(title)	81	print(title)
84	```	82	```
85		83
86	Or you can parse a local pdf:	84	Or you can parse a local pdf:
87		85
88	```python	86	```python
89	from aitrika.engine.aitrika import LocalAItrika	87	from aitrika.engine.aitrika import LocalAItrika
90	aitrika_engine = LocalAItrika(pdf_path = pdf_path)	88	aitrika_engine = LocalAItrika(pdf_path = pdf_path)
91	title = aitrika_engine.get_title()	89	title = aitrika_engine.get_title()
92	print(title)	90	print(title)
93	```	91	```
94		92
95	```	93	```
96	Breast cancer genes: beyond BRCA1 and BRCA2.	94	Breast cancer genes: beyond BRCA1 and BRCA2.
97	```	95	```
98		96
99	You can get other informations, like the associations between genes and diseases:	97	You can get other informations, like the associations between genes and diseases:
100		98
101	```python	99	```python
102	associations = aitrika_engine.get_associations()	100	associations = aitrika_engine.get_associations()
103	```	101	```
104		102
105	```	103	```
106	[	104	[
107	{	105	{
108	"gene": "BRIP1",	106	"gene": "BRIP1",
109	"disease": "Breast Neoplasms"	107	"disease": "Breast Neoplasms"
110	},	108	},
111	{	109	{
112	"gene": "PTEN",	110	"gene": "PTEN",
113	"disease": "Breast Neoplasms"	111	"disease": "Breast Neoplasms"
114	},	112	},
115	{	113	{
116	"gene": "CHEK2",	114	"gene": "CHEK2",
117	"disease": "Breast Neoplasms"	115	"disease": "Breast Neoplasms"
118	},	116	},
119	]	117	]
120	...	118	...
121	```	119	```
122		120
123	Or you can get a nice formatted DataFrame:	121	Or you can get a nice formatted DataFrame:
124		122
125	```python	123	```python
126	associations = aitrika_engine.associations(dataframe = True)	124	associations = aitrika_engine.associations(dataframe = True)
127	```	125	```
128		126
129	```	127	```
130	gene disease	128	gene disease
131	0 BRIP1 Breast Neoplasms	129	0 BRIP1 Breast Neoplasms
132	1 PTEN Breast Neoplasms	130	1 PTEN Breast Neoplasms
133	2 CHEK2 Breast Neoplasms	131	2 CHEK2 Breast Neoplasms
134	...	132	...
135	```	133	```
136		134
137	With the power of RAG, you can query your document:	135	With the power of RAG, you can query your document:
138		136
139	```python	137	```python
140	## Prepare the documents	138	## Prepare the documents
141	documents = generate_documents(content=abstract)	139	documents = generate_documents(content=abstract)
142		140
143	## Set the LLM	141	## Set the LLM
144	llm = GroqLLM(documents=documents, api_key=os.getenv("GROQ_API_KEY"))	142	llm = GroqLLM(documents=documents, api_key=os.getenv("GROQ_API_KEY"))
145		143
146	## Query your document	144	## Query your document
147	query = "Is BRCA1 associated with breast cancer?"	145	query = "Is BRCA1 associated with breast cancer?"
148	print(llm.query(query=query))	146	print(llm.query(query=query))
149	```	147	```
150		148
151	```	149	```
152	The provided text suggests that BRCA1 is associated with breast cancer, as it is listed among the high-penetrance genes identified in family linkage studies as responsible for inherited syndromes of breast cancer.	150	The provided text suggests that BRCA1 is associated with breast cancer, as it is listed among the high-penetrance genes identified in family linkage studies as responsible for inherited syndromes of breast cancer.
153	```	151	```
154		152
155	Or you can extract other informations:	153	Or you can extract other informations:
156		154
157	```python	155	```python
158	results = engine.extract_results(llm=llm)	156	results = engine.extract_results(llm=llm)
159	print(results)	157	print(results)
160	```	158	```
161		159
162	```	160	```
163	RESULTS	161	RESULTS
164		162
165	- High-penetrance genes - BRCA1, BRCA2, PTEN, TP53 - responsible for inherited syndromes	163	- High-penetrance genes - BRCA1, BRCA2, PTEN, TP53 - responsible for inherited syndromes
166	- Moderate-penetrance genes - CHEK2, ATM, BRIP1, PALB2, RAD51C - associated with moderate BC risk	164	- Moderate-penetrance genes - CHEK2, ATM, BRIP1, PALB2, RAD51C - associated with moderate BC risk
167	- Low-penetrance alleles - common alleles - associated with slightly increased or decreased risk of BC	165	- Low-penetrance alleles - common alleles - associated with slightly increased or decreased risk of BC
168	- Current clinical practice - high-penetrance genes - widely used	166	- Current clinical practice - high-penetrance genes - widely used
169	- Future prospect - all familial breast cancer genes - to be included in genetic test	167	- Future prospect - all familial breast cancer genes - to be included in genetic test
170	- Research need - clinical management - of moderate and low-risk variants	168	- Research need - clinical management - of moderate and low-risk variants
171	```	169	```
172		170
173	## 🚀 Run the API	171	## 🚀 Run the API
174		172
175	To run the AItrika API, follow these steps:	173	To run the AItrika API, follow these steps:
176		174
177	1. Ensure you have set up your environment and installed all dependencies as described in the Installation section.	175	1. Ensure you have set up your environment and installed all dependencies as described in the Installation section.
178		176
179	2. Run the API server using the following command:	177	2. Run the API server using the following command:
180		178
181	```bash	179	```bash
182	python api.py	180	python api.py
183	```	181	```
184		182
185	The API will start running on http://0.0.0.0:8000. You can now make requests to the various endpoints:	183	The API will start running on http://0.0.0.0:8000. You can now make requests to the various endpoints:
186		184
187	- /associations: Get associations from a PubMed article	185	- /associations: Get associations from a PubMed article
188	- /abstract: Get abstract of a PubMed article	186	- /abstract: Get abstract of a PubMed article
189	- /query: Query a PubMed article	187	- /query: Query a PubMed article
190	- /results: Get results from a PubMed article	188	- /results: Get results from a PubMed article
191	- /participants: Get number of participants from a PubMed article	189	- /participants: Get number of participants from a PubMed article
192	- /outcomes: Get outcomes from a PubMed article	190	- /outcomes: Get outcomes from a PubMed article
193		191
194	You can use tools like curl, Postman, or any HTTP client to interact with the API. For example:	192	You can use tools like curl, Postman, or any HTTP client to interact with the API. For example:
195		193
196	```bash	194	```bash
197	curl -X POST "http://localhost:8000/abstract" -H "Content-Type: application/json" -d '{"pubmed_id": 12345678}'	195	curl -X POST "http://localhost:8000/abstract" -H "Content-Type: application/json" -d '{"pubmed_id": 12345678}'
198	```	196	```
199		197
200	The API documentation is automatically generated and saved to <code>docs/api-reference/openapi.json</code>.	198	The API documentation is automatically generated and saved to <code>docs/api-reference/openapi.json</code>.
201	You can use this file with tools like Swagger UI for a more interactive API exploration experience.	199	You can use this file with tools like Swagger UI for a more interactive API exploration experience.
202		200
203	## Support the Project	201	## Support the Project
204		202
205	If you find this project useful, please consider supporting it:	203	If you find this project useful, please consider supporting it:
206		204
207	- 🌟 Star the project on GitHub	205	- 🌟 Star the project on GitHub
208	- 🐛 Report bugs or suggest new features	206	- 🐛 Report bugs or suggest new features
209	- 🤝 Contribute with pull requests	207	- 🤝 Contribute with pull requests
210	- ☕️ [Buy me a coffee](https://www.buymeacoffee.com/dsupertramp) or consider a sponsor.	208	- ☕️ [Buy me a coffee](https://www.buymeacoffee.com/dsupertramp) or consider a sponsor.
211		209
212	### Commercial / Business use	210	### Commercial / Business use
213		211
214	If you're using this project in a business or commercial context, please [contact me](salvatoredanilopalumbo@gmail.com).	212	If you're using this project in a business or commercial context, please [contact me](salvatoredanilopalumbo@gmail.com).
215		213
216	I'm available for consulting, custom development, or commercial licensing.	214	I'm available for consulting, custom development, or commercial licensing.
217		215
218	Your support helps keep this project active and continuously improving. Thank you!	216	Your support helps keep this project active and continuously improving. Thank you!
219		217
220	## License	218	## License
221		219
222	AItrika is licensed under the Apache 2.0 License. See the LICENSE file for more details.	220	AItrika is licensed under the Apache 2.0 License. See the LICENSE file for more details.
223		221
224	## Star History	222	## Star History
225		223
226	[![Star History Chart](https://api.star-history.com/svg?repos=dSupertramp/AItrika&type=Date)](https://star-history.com/#dSupertramp/AItrika&Date)	224	[![Star History Chart](https://api.star-history.com/svg?repos=dSupertramp/AItrika&type=Date)](https://star-history.com/#dSupertramp/AItrika&Date)