Molecular de-novo design through deep reinforcement learning

TLDR

The study proposes a sequence‑based generative model tuned via augmented episodic likelihood to design de‑novo molecules with user‑specified desirable properties. The model is trained to generate molecules under constraints such as absence of sulfur, analogues of a query compound, or predicted activity against a target, demonstrating tasks like scaffold hopping and library expansion. When optimized for dopamine receptor D2 activity, the model produced over 95 % of molecules predicted active, including experimentally verified actives not present in the training data.

Abstract

This work introduces a method to tune a sequence-based generative model for molecular de novo design that through augmented episodic likelihood can learn to generate structures with certain specified desirable properties. We demonstrate how this model can execute a range of tasks such as generating analogues to a query structure and generating compounds predicted to be active against a biological target. As a proof of principle, the model is first trained to generate molecules that do not contain sulphur. As a second example, the model is trained to generate analogues to the drug Celecoxib, a technique that could be used for scaffold hopping or library expansion starting from a single molecule. Finally, when tuning the model towards generating compounds predicted to be active against the dopamine receptor type 2, the model generates structures of which more than 95% are predicted to be active, including experimentally confirmed actives that have not been included in either the generative model nor the activity prediction model.