Skip to content

Evaluation of Reinforcement Learning in Transformer-based Molecular Design

License

Notifications You must be signed in to change notification settings

MolecularAI/transformer_rl

Repository files navigation

Evaluation of Reinforcement Learning in Transformer-based Molecular Design

Description

This repository holds the code for Evaluation of Reinforcement Learning in Transformer-based Molecular Design. It was implemented based on the Reinvent3 framework.

Installation

  1. Install Conda

  2. Clone this Git repository

  3. Open a shell, and go to the repository and create the Conda environment:

     $ conda env create -f reinvent.yml
    
  4. Activate the environment:

     $ conda activate reinvent
    

Usage

In support directory, there are the PubChem prior, ChEMBL prior, DRD2 predictive model used in the manuscript. Additionally, there is an example input configuration example.json for running the code,

python input.py support/example.json

About

Evaluation of Reinforcement Learning in Transformer-based Molecular Design

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages