liyuesen wxyedward commited on
Commit
9c03bc4
β€’
1 Parent(s): 5c3b926

Update README.md (#6)

Browse files

- Update README.md (02ef01255cc09c688bf7ecd4cd4bd35ed63545b1)


Co-authored-by: XJTU_WangXiangyu <[email protected]>

Files changed (1) hide show
  1. README.md +36 -8
README.md CHANGED
@@ -1,5 +1,5 @@
1
  ---
2
- license: artistic-2.0
3
  tags:
4
  - chemistry
5
  - biology
@@ -8,8 +8,14 @@ tags:
8
  ---
9
  # DrugGPT
10
  A generative drug design model based on GPT2.
11
- <img src="https://img.shields.io/badge/license-Artistic%20License%202.0-green"><img src="https://img.shields.io/badge/python-3.7-blue"><img src="https://img.shields.io/github/stars/LIYUESEN/druggpt?style=social">
12
- ## Deployment
 
 
 
 
 
 
13
  1. Clone
14
  ```shell
15
  git clone https://github.com/LIYUESEN/druggpt.git
@@ -27,8 +33,21 @@ A generative drug design model based on GPT2.
27
  pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
28
  conda install -c openbabel openbabel
29
  ```
30
- ## Example usage
31
- Run the script with the desired arguments, such as the protein sequence, ligand prompt, number of molecules to generate, and output directory.
 
 
 
 
 
 
 
 
 
 
 
 
 
32
  - If you want to input a protein FASTA file
33
  ```shell
34
  python drug_generator.py -f bcl2.fasta -n 50
@@ -42,9 +61,18 @@ Run the script with the desired arguments, such as the protein sequence, ligand
42
  ```shell
43
  python drug_generator.py -f bcl2.fasta -l COc1ccc(cc1)C(=O) -n 50
44
  ```
 
45
  - Note: If you are running in a Linux environment, you need to enclose the ligand's prompt with single quotes ('').
46
  ```shell
47
  python drug_generator.py -f bcl2.fasta -l 'COc1ccc(cc1)C(=O)' -n 50
48
- ```
49
- ## License
50
- [Artistic License 2.0](https://opensource.org/license/artistic-license-2-0-php/)
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: gpl-3.0
3
  tags:
4
  - chemistry
5
  - biology
 
8
  ---
9
  # DrugGPT
10
  A generative drug design model based on GPT2.
11
+ <img src="https://img.shields.io/github/license/LIYUESEN/druggpt"><img src="https://img.shields.io/badge/python-3.7-blue"><img src="https://img.shields.io/github/stars/LIYUESEN/druggpt?style=social">
12
+ ## 🚩 Introduction
13
+ DrugGPT is a generative pharmaceutical strategy based on GPT structure, which aims to bring innovation to drug design by using natural language processing technique.
14
+
15
+ This project applies the GPT model to the exploration of chemical space to discover new molecules with potential binding abilities for specific proteins.
16
+
17
+ DrugGPT provides a fast and efficient method for the generation of drug candidate molecules by training on up to 1.8 million protein-ligand binding data.
18
+ ## πŸ“₯ Deployment
19
  1. Clone
20
  ```shell
21
  git clone https://github.com/LIYUESEN/druggpt.git
 
33
  pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
34
  conda install -c openbabel openbabel
35
  ```
36
+ ## πŸ— How to use
37
+ Use [drug_generator.py](https://github.com/LIYUESEN/druggpt/blob/main/drug_generator.py)
38
+
39
+ Required parameters:
40
+ - `-p` | `--pro_seq`: Input a protein amino acid sequence.
41
+ - `-f` | `--fasta`: Input a FASTA file.
42
+
43
+ > Only one of -p and -f should be specified.
44
+ - `-l` | `--ligand_prompt`: Input a ligand prompt.
45
+ - `-e` | `--empty_input`: Enable directly generate mode.
46
+ - `-n` | `--number`: At least how many molecules will be generated.
47
+ - `-d` | `--device`: Hardware device to use. Default is 'cuda'.
48
+ - `-o` | `--output`: Output directory for generated molecules. Default is './ligand_output/'.
49
+ - `-b` | `--batch_size`: How many molecules will be generated per batch. Try to reduce this value if you have low RAM. Default is 32.
50
+ ## πŸ”¬ Example usage
51
  - If you want to input a protein FASTA file
52
  ```shell
53
  python drug_generator.py -f bcl2.fasta -n 50
 
61
  ```shell
62
  python drug_generator.py -f bcl2.fasta -l COc1ccc(cc1)C(=O) -n 50
63
  ```
64
+
65
  - Note: If you are running in a Linux environment, you need to enclose the ligand's prompt with single quotes ('').
66
  ```shell
67
  python drug_generator.py -f bcl2.fasta -l 'COc1ccc(cc1)C(=O)' -n 50
68
+ ```
69
+ ## πŸ“ How to reference this work
70
+ DrugGPT: A GPT-based Strategy for Designing Potential Ligands Targeting Specific Proteins
71
+
72
+ Yuesen Li, Chengyi Gao, Xin Song, Xiangyu Wang, Yungang Xu, Suxia Han
73
+
74
+ bioRxiv 2023.06.29.543848; doi: [https://doi.org/10.1101/2023.06.29.543848](https://doi.org/10.1101/2023.06.29.543848)
75
+
76
+ [![DOI](https://img.shields.io/badge/DOI-10.1101/2023.06.29.543848-blue)](https://doi.org/10.1101/2023.06.29.543848)
77
+ ## βš– License
78
+ [GNU General Public License v3.0](https://www.gnu.org/licenses/gpl-3.0.html)