Submitted:
19 November 2020
Posted:
20 November 2020
You are already at the latest version
Abstract
In this paper we propose the generation of synthetic small and more sophisticated molecule structures that optimize the binding affinity to a target (ASYNT-GAN). To achieve this we leverage on three important achievements in A.I.: Attention, Deep Learning on Graphs and Generative Adversarial Networks. Similar to text generation based on parts of text we are able to generate a molecule architecture based on an existing target. By adopting this approach, we propose a novel way of searching for existing compounds that are suitable candidates. Similar to question and answer Natural Language solutions we are able to find drugs with highest relevance to a target. We are able to identify substructures of the molecular structure that are the most suitable for binding. In addition, we are proposing a novel way of generating the molecule in 3D space in such a way that the binding is optimized. We show that we are able to generate compound structures and protein structures that are optimised for binding to a target.
Keywords:
drug discovery
; artificial intelligence
; protein discovery
; binding prediction
; synthetic molecule generation
; synthetic drug
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
