Implementation of the method proposed in the paper "Evolutionary multi-objective molecule optimization in implicit chemical space" by Xin Xia, Yansen Su, Yiping Liu, Chunhou Zheng, Xiangxiang Zeng.1
- cddd Download the pre-trained CDDD model using the bash script:
./download_default_model.sh
- python=3.6
- rdkit
- pytorch=1.4.0
- cudatoolkit=10.0
- tensorboardX
- pip>=19.1,<20.3
- pip:
- molsets
- cddd
cd MOMO
pip install .
For QED and Similarity optimization Task 1, please run python MOMO_task1.py For Plogp and Similarity optimization Task 2, please run python MOMO_task2.py For QED, Drd2 and Similarity optimization Task 3, please run python MOMO_task3.py
The output results of molecules are summarized in smi_pro_tuple, and further save in .csv file.
The fitness function can wrap any function that has following properties:
- Takes a RDKit mol object as input and returns a number as score.
If you use this work, please cite the following:
@article{xia2022molecule,
title={Molecule optimization via multi-objective evolutionary in implicit chemical space},
author={Xia, Xin and Su, Yansen and Zheng, Chunhou and Zeng, Xiangxiang},
journal={arXiv preprint arXiv:2212.08826},
year={2022}
}