Deep Evolutionary Learning for Molecular Design

Yifeng Li, Hsu Kiang Ooi, Alain Tchagang

Keywords: Deep Evolutionary Learning, Fragment-Based Drug Design, Deep Generative Model, Drug Design, Multi-objective Optimization

Abstract: In this paper, we propose a deep evolutionary learning (DEL) process that integrates fragment-based deep generative model and multi-objective evolutionary computation for molecular design. Our approach enables (1) evolutionary operations in the latent space of the generative model, rather than the structural space, to generate novel promising molecular structures for the next evolutionary generation, and (2) generative model fine-tuning using newly generated high-quality samples. Thus, DEL implements a data-model co-evolution concept which improves both sample population and generative model learning. Experiments on two public datasets indicate that sample population obtained by DEL exhibits improved property distributions, and dominates samples generated by multi-objective Bayesian optimization algorithms.

One-sentence Summary: Population of new molecules designed by our novel deep evolutionary learning process exhibit improved values of properties in comparison with original training molecules.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Reviewed Version (pdf): https://openreview.net/references/pdf?id=fHMPX_EmRd