A PyTorch implementation of “GxVAEs: Two Joint VAEs Generate Hit Molecules from Gene Expression Profiles“. The paper has been accepted by AAAI 2024.

This implementation was developed by Chen Li (li.chen.z2@a.mail.nagoya-u.ac.jp) and Yoshihiro Yamanishi (yamanishi@i.nagoya-u.ac.jp), affiliated with the Department of Complex Systems Science at the Graduate School of Informatics, Nagoya University, Japan, at the time of release.

GxVAEs aim to

• generate hit-like molecules from gene expression profiles.
• generate therapeutic molecules from patients’ disease profiles.

Execute the following command:

$ conda env create -n gxvae_env -f gxvaes_env.yml
$ source activate gxvaes_env

• The datasets Folder
  • LINCS/mcf7.csv: The training and validation datasets, which consist of gene expression profiles of the MCF7 cell line treated with 13,755 molecules, were used.
  • tools floder
• main.py:: Define the main function for training the ProfileVAE and MolVAE models.
• ProfileVAE.py: Defines the ProfileVAE model for extracting gene expression profile features.
• train_gene_vae.py: Code for training the ProfileVAE model.
• MolVAE.py: Defines the MolVAE model to generate SMILES strings with extracted gene features.
• train_smiles_vae.py: Code for training the MolVAE model.
• utils.py: Defines other functions used in GxVAEs.

• STEP 1: Pretrain ProfileVAE:

$ python main.py --train_gene_vae

• STEP 2: Test the trained ProfileVAE:

$ python main.py --test_gene_vae

• STEP 3: Train MolVAE:

$ python main.py --train_smiles_vae

• STEP 4: Test the trained MolVAE:

$ python main.py --test_smiles_vae

• STEP 5: Generate molecules for the 10 ligands using GxVAEs

$ python main.py --generation

• STEP 6: Calculate Tanimoto similarity between a source ligand and generated SMILES strings:

$ python main.py --calculate_tanimoto --protein_name ***

    Note that '***' indicates a protein name, such as 'AKT1'.

If you have any questions, please feel free to contact Chen Li at li.chen.z2@a.mail.nagoya-u.ac.jp.

C. Li and Y. Yamanishi (2024). GxVAEs： Two Joint VAEs Generate Hit Molecules from Gene Expression Profiles.

BibTeX format:

@inproceedings{li2024gxvaes,
title={GxVAEs： Two Joint VAEs Generate Hit Molecules from Gene Expression Profiles},
author={Li, Chen and Yamanishi, Yoshihiro},
booktitle={Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI 2024)},
year={2024}
}