REINVENT-Transformer: Molecular De Novo Design through Transformer-based Reinforcement Learning

Pengcheng Xu, Tianfan Fu, Wenhao Gao, Jimeng Sun

Keywords: Molecular Optimization, Transformer, Reinforcement Learning

TL;DR: REINVENT-Transformer improves molecular design by using Transformer-based models, outperforming RNN-based methods in generating effective compounds for various biological targets. Abstract:

Abstract: In this work, we introduce a method: REINVENT-Transformer to fine-tune a Transformer-based generative model for molecular de novo design. Leveraging the superior sequence learning capacity of Transformers over Recurrent Neural Networks (RNNs), our model can generate molecular structures with desired properties effectively. In contrast to the traditional RNN-based models, our proposed method exhibits superior performance in generating compounds predicted to be active against various biological targets, capturing long-term dependencies in the molecular structure sequence. The model's efficacy is demonstrated across numerous tasks, including generating analogues to a query structure and producing compounds with particular attributes, outperforming the baseline RNN-based methods. Our approach can be used for scaffold hopping, library expansion starting from a single molecule and generating compounds with high predicted activity against biological targets.

Submission Number: 41