Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/josiehong/3dmolcsp
[Analytical Chemistry] Enhanced Structure-Based Prediction of Chiral Stationary Phases for Chromatographic Enantioseparation from 3D Molecular Conformations
https://github.com/josiehong/3dmolcsp
analytical-chemistry chromatography molecule
Last synced: 27 days ago
JSON representation
[Analytical Chemistry] Enhanced Structure-Based Prediction of Chiral Stationary Phases for Chromatographic Enantioseparation from 3D Molecular Conformations
- Host: GitHub
- URL: https://github.com/josiehong/3dmolcsp
- Owner: JosieHong
- License: other
- Created: 2023-04-29T18:51:58.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-02-04T20:04:03.000Z (9 months ago)
- Last Synced: 2024-09-22T13:32:32.327Z (about 1 month ago)
- Topics: analytical-chemistry, chromatography, molecule
- Language: Python
- Homepage:
- Size: 15.8 MB
- Stars: 5
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# 3DMolCSP
[![CC BY-NC-SA 4.0][cc-by-nc-sa-shield]][cc-by-nc-sa] (free for academic use)
Enhanced Structure-Based Prediction of Chiral Stationary Phases for Chromatographic Enantioseparation from 3D Molecular Conformations
Yuhui Hong, Christopher J. Welch, Patrick Piras, and Haixu Tang*
This work introduces 3D molecular representation learning to enhance the prediction of enantioselectivity of Chiral Stationary Phases (CSPs) and elution order of enantiomers in the field of chromatography, which can assist in the selection of CSPs for practical applications. Our [[paper]](https://pubs.acs.org/doi/full/10.1021/acs.analchem.3c04028) published on Analytical Chemistry provides further details.
## Set up
```bash
# RDKit 2022.03.5
# https://www.rdkit.org/docs/GettingStartedInPython.html
conda create -c rdkit -n rdkit
conda activate
# please use rdkit >= 2021.03.4
# https://github.com/rdkit/rdkit/pull/4272
# Pytorch 1.13.0
# Please choose the proper cuda version from their official website:
# https://pytorch.org/get-started/previous-versions/
conda install pytorch==1.13.0 torchvision==0.14.0 torchaudio==0.13.0 pytorch-cuda=11.7 -c pytorch -c nvidia
conda install -c conda-forge tensorboard
pip install lxml tqdm pandas pyteomics PyYAML scikit-learn
```
## Data preprocessing
### Demo set (Chirobiotic V)
```bash
# The demo dataset is already cleaned.
# generate 3D conformations
python ./preprocess/gen_conformers.py --path ./data/demo/chirobiotic_v.sdf --conf_type etkdg
```
### ChirBase
```bash
# preprocessing
python ./preprocess/preprocess_chirbase.py \
--input ./data/ChirBase/chirbase.sdf \
--output ./data/ChirBase/chirbase_clean.sdf \
--csp_setting ./preprocess/chirality_stationary_phase_list.csv
# generate 3D conformations
python ./preprocess/gen_conformers.py --path ./data/ChirBase/chirbase_clean.sdf --conf_type etkdg
# (optional) OMEGA conformations are available
# OMEGA algorithm requires a license. Please get the license here: https://www.eyesopen.com/omega
# python ./preprocess/gen_conformers.py --path ./data/ChirBase/chirbase_clean.sdf --conf_type omega --license
# (optional) randomly split training and validation set for Exp3
python ./preprocess/random_split_sdf.py \
--input ./data/ChirBase/chirbase_clean_etkdg.sdf \
--output_train ./data/ChirBase/chirbase_clean_etkdg_train.sdf \
--output_test ./data/ChirBase/chirbase_clean_etkdg_test.sdf
# python ./preprocess/random_split_sdf.py \
# --input ./data/ChirBase/chirbase_clean_omega.sdf \
# --output_train ./data/ChirBase/chirbase_clean_omega_train.sdf \
# --output_test ./data/ChirBase/chirbase_clean_omega_test.sdf
```
### CMRT
```bash
# preprocessing
python ./preprocess/preprocess_cmrt.py \
--input ./data/CMRT/cmrt_all_column.csv \
--output ./data/CMRT/cmrt_clean.sdf \
--csp_setting ./preprocess/chirality_stationary_phase_list.csv
# generate 3D conformations
python ./preprocess/gen_conformers.py --path ./data/CMRT/cmrt_clean.sdf --conf_type etkdg
```
## Experiments
### Exp1: Demo
1. [Preprocess demo dataset](#demo-set-chirobiotic-v)
2. Five-fold cross-validation
```bash
# training from scratch
python main_chir_kfold.py --config ./configs/molnet_train_demo.yaml --k_fold 5 --csp_no 3 \
--log_dir ./logs/molnet_chirality/ \
--checkpoint ./check_point/demo_sc.pt \
--result_path ./demo/demo_sc.csv \
--device 1
# training from pretrained model
python main_chir_kfold.py --config ./configs/molnet_train_demo.yaml --k_fold 5 --csp_no 3 \
--log_dir ./logs/molnet_chirality/ \
--resume_path ./check_point/molnetv2_qtof_etkdgv3.pt \
--transfer \
--checkpoint ./check_point/demo_tl.pt \
--result_path ./demo/demo_tl.csv \
--device 1
```
Results on demo dataset:
```bash
# --------------- Final Results of 3DMolCSP-SC --------------- #
fold_0: acc: 0.81, auc: 0.95
fold_1: acc: 0.90, auc: 0.99
fold_2: acc: 0.73, auc: 0.87
fold_3: acc: 0.81, auc: 0.96
fold_4: acc: 0.84, auc: 0.98
mean acc: 0.82, mean auc: 0.95
# --------------- Final Results of 3DMolCSP-TL --------------- #
fold_0: acc: 0.80, auc: 0.94
fold_1: acc: 0.91, auc: 0.99
fold_2: acc: 0.81, auc: 0.97
fold_3: acc: 0.79, auc: 0.96
fold_4: acc: 0.80, auc: 0.97
mean acc: 0.82, mean auc: 0.97
```
### Exp2: Five-fold cross-validation on ChirBase
1. [Preprocess ChirBase](#chirbase)
2. Five-fold cross-validation
```bash
# training from scratch
nohup bash ./experiments/train_chir_etkdg_5fold.sh > molnet_chir_etkdg_5fold.out
# traning from pre-trained model
nohup bash ./experiments/train_chir_etkdg_5fold_tl.sh > molnet_chir_etkdg_5fold_tl.out
```
### Exp3: Training on ChirBase and testing on CMRT
1. Training (using all data)
```bash
# traning from pre-trained model
nohup bash ./experiments/train_chir_etkdg_tl.sh > molnet_chir_etkdg_tl.out
```
2. [Preprocess CMRT](#cmrt)
3. infer on CMRT
```bash
nohup bash ./experiments/infer_cmrt_etkdg_tl.sh > molnet_cmrt_etkdg_tl_infer.out
```
### Exp4: Elution order prediction
```bash
nohup bash ./experiments/train_chir_etkdg_eo.sh > train_chir_etkdg_eo.out
```
---
This work is licensed under a
[Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License][cc-by-nc-sa].
[![CC BY-NC-SA 4.0][cc-by-nc-sa-image]][cc-by-nc-sa]
[cc-by-nc-sa]: http://creativecommons.org/licenses/by-nc-sa/4.0/
[cc-by-nc-sa-image]: https://licensebuttons.net/l/by-nc-sa/4.0/88x31.png
[cc-by-nc-sa-shield]: https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg