This repository contains the implementation code for paper PepScaf: Harnessing Machine Learning with In Vitro Selection toward De Novo Macrocyclic Peptides against IL-17C/IL-17RE Interaction.
In this study, we first conducted the in vitro selection of macrocyclic peptides against interleukin-17C (IL-17C) using a primary library (17C-Lib1), which gave us a macrocyclic peptide ligand capable of inhibiting IL-17C/IL-17RE interaction with the
The code was test on GPU 3060 with
python=3.8pytorch==1.12.1cudatoolkit=11.3
We use conda to install the dependencies for PepScaf from the provided environment.yml file, which can give you the exact python environment we run the code for the paper:
NOTE: we also highly recommend using mamba instead of vanilla conda for managing your conda environments. Mamba is a drop-in replacement for conda that is:
- Faster at solving environments (>10x in my experience)
- Better at resolving conflicts
- More informative when something goes wrong.
git clone https://github.com/hongliangduan/PepScaf.git
cd PepScaf
mamba env create -f environment.yml
pip install sci-ztools==0.1.0
mamba activate pepscafpython bin/preprocessing.py # preprocessing for raw data
python bin/cluster.py # get cluster to get targetsbash train.sh
bash eval.shbash get_attn.shRun notebooks/Score.ipynb to get position scores.
bash mcts.sh # run mcts to get the scaffold- The implementation of Pep-BERT is partly inspired by A transformer-based model to predict peptide–HLA class I binding and optimize mutated peptides for vaccine design and MolSearch: Search-based multi-objective molecular generation and property optimization.
- The Pep-BERT was build based fon pytorchic-bert, which is a re-implementation of Google BERT model in Pytorch.
- The code was formattered by Black.
- The building of MCTS parts refered to int8's mcts.
- Weblog
- The visualization of attention was modified from Pytorch Community
- The CD-HIT tool and Biopython was used for clustering sequence-based macrocyclic peptides.
If you find the idea or code useful for your research, please cite our paper:
@article{doi:10.1021/acs.jmedchem.3c00627,
author = {Zhai, Silong and Tan, Yahong and Zhang, Chengyun and Hipolito, Christopher John and Song, Lulu and Zhu, Cheng and Zhang, Youming and Duan, Hongliang and Yin, Yizhen},
title = {PepScaf: Harnessing Machine Learning with In Vitro Selection toward De Novo Macrocyclic Peptides against IL-17C/IL-17RE Interaction},
journal = {Journal of Medicinal Chemistry},
volume = {0},
number = {0},
pages = {null},
year = {0},
doi = {10.1021/acs.jmedchem.3c00627},
note ={PMID: 37480587},
URL = {
https://doi.org/10.1021/acs.jmedchem.3c00627
},
eprint = {
https://doi.org/10.1021/acs.jmedchem.3c00627
}
Please contact hduan@zjut.edu.cn if you have any question. Enjoy!