fate1997 / TSTiNet

A noval transition state theory - inspired neural network for the prediction of deep eutectic solvents (DES)

Transition state theory-inspired neural network for estimation of the viscosity of deep eutectic solvents

This is a implentation of our paper "Transition state theory-inspired neural network for estimation of the viscosity of deep eutectic solvents":

Requirements

• scikit-learn == 0.24.2
• pytorch == 1.9.0+cu111
• numpy == 1.18.5
• lightgbm == 3.2.1
• pandas == 1.2.4

Model structure

How to use

If you want to use our model to predict the viscosity of specified deep eutectic solvents (DES), you can follow these steps:

1. git clone https://github.com/fate1997/TSTiNet
2. cd TSTiNet/prediction
3. open the "input.xlsx" file and fill the rows.
   (ATTNTION: DO NOT DELETE THE "example" ROW)
4. python predict.py

If you want to train a new model, you can just run the code:

cd model && python TSTiNet-mixed.py

Cite

If you use TSTiNet in your research, please cite:

@article{doi:10.1021/acscentsci.2c00157,
author = {Yu, Liu-Ying and Ren, Gao-Peng and Hou, Xiao-Jing and Wu, Ke-Jun and He, Yuchen},
title = {Transition State Theory-Inspired Neural Network for Estimating the Viscosity of Deep Eutectic Solvents},
journal = {ACS Central Science},
volume = {8},
number = {7},
pages = {983-995},
year = {2022},
doi = {10.1021/acscentsci.2c00157}}