We Should at Least Be Able to Design Molecules That Dock Well

TL;DR

This paper introduces a docking-based benchmark for evaluating molecule generation models in drug discovery, highlighting current limitations and providing a simplified version to facilitate progress in designing molecules with high binding affinity.

Contribution

It proposes a new benchmark based on docking scores, reveals the limitations of existing graph-based models, and offers a simplified scoring version for future research.

Findings

Popular models struggle to generate high-scoring molecules

Current models have limitations in realistic drug design scenarios

A simplified benchmark can be partially solved by existing models

Abstract

Designing compounds with desired properties is a key element of the drug discovery process. However, measuring progress in the field has been challenging due to the lack of realistic retrospective benchmarks, and the large cost of prospective validation. To close this gap, we propose a benchmark based on docking, a popular computational method for assessing molecule binding to a protein. Concretely, the goal is to generate drug-like molecules that are scored highly by SMINA, a popular docking software. We observe that popular graph-based generative models fail to generate molecules with a high docking score when trained using a realistically sized training set. This suggests a limitation of the current incarnation of models for de novo drug design. Finally, we propose a simplified version of the benchmark based on a simpler scoring function, and show that the tested models are able to partially solve it. We release the benchmark as an easy to use package available at https://github.com/cieplinski-tobiasz/smina-docking-benchmark. We hope that our benchmark will serve as a stepping stone towards the goal of automatically generating promising drug candidates.