GuacaMol: Benchmarking Models for de Novo Molecular Design

TLDR

De novo design aims to generate molecules with desired property profiles through virtual design‑make‑test cycles, yet recent neural generative models lack consistent profiling and comparative evaluation. This work introduces GuacaMol, an evaluation framework that standardizes the assessment of both classical and neural de novo molecular design models. GuacaMol’s benchmark suite evaluates models on fidelity to training property distributions, novelty generation, chemical space exploration and exploitation, and single‑ and multi‑objective optimization tasks. The open‑source Python implementation and leaderboard are available at https://benevolent.ai/guacamol.

Abstract

De novo design seeks to generate molecules with required property profiles by virtual design-make-test cycles. With the emergence of deep learning and neural generative models in many application areas, models for molecular design based on neural networks appeared recently and show promising results. However, the new models have not been profiled on consistent tasks, and comparative studies to well-established algorithms have only seldom been performed. To standardize the assessment of both classical and neural models for de novo molecular design, we propose an evaluation framework, GuacaMol, based on a suite of standardized benchmarks. The benchmark tasks encompass measuring the fidelity of the models to reproduce the property distribution of the training sets, the ability to generate novel molecules, the exploration and exploitation of chemical space, and a variety of single and multiobjective optimization tasks. The benchmarking open-source Python code and a leaderboard can be found on https://benevolent.ai/guacamol .