Optimization of binding affinities in chemical space with generative pre-trained transformer and deep reinforcement learning -- source data

This repository contains the data for SGPT-RL, a tool for chemical design using generative pre-trained transformer and deep reinforcement learning. Through employing GPT model as the policy network, SGPT-RL can learn scaffold patterns in exploring the chemical space. Publication: TBA Files included in this release: sgpt-rl.png: the workflow of SGPT-RL. data.zip: raw datasets and pretrained models. Chemical_GPT_SI.pdf: supplementary tables and figures. Source code repository: https://github.com/charlesxu90/sgpt