Bridging the gap between target-based and cell-based drug discovery with a graph generative multi-task model

TL;DR

This paper introduces a graph multi-task deep learning model that predicts compounds with both target inhibition and cell activity, and uses reinforcement learning to generate novel multi-property drugs, bridging in vitro and in vivo drug discovery gaps.

Contribution

It presents a novel multi-task graph neural network for dual-property prediction and a reinforcement learning approach for generating multi-effective compounds.

Findings

MATIC model outperforms traditional methods in SARS-CoV-2 compound screening.

Learned features differ between target inhibition and cell activity tasks.

Generated compounds show potential for dual efficacy in vitro and in vivo.

Abstract

Drug discovery is vitally important for protecting human against disease. Target-based screening is one of the most popular methods to develop new drugs in the past several decades. This method efficiently screens candidate drugs inhibiting target protein in vitro, but it often fails due to inadequate activity of the selected drugs in vivo. Accurate computational methods are needed to bridge this gap. Here, we propose a novel graph multi task deep learning model to identify compounds carrying both target inhibitory and cell active (MATIC) properties. On a carefully curated SARS-CoV-2 dataset, the proposed MATIC model shows advantages comparing with traditional method in screening effective compounds in vivo. Next, we explored the model interpretability and found that the learned features for target inhibition (in vitro) or cell active (in vivo) tasks are different with molecular property correlations and atom functional attentions. Based on these findings, we utilized a monte carlo based reinforcement learning generative model to generate novel multi-property compounds with both in vitro and in vivo efficacy, thus bridging the gap between target-based and cell-based drug discovery.