Deep generative models are at the forefront of deep learning research, because of their potential in data-efficient learning, and model-based reinforcement learning. Those are also becoming increasingly popular in tackling the computational de novo drug design problem, which requires exploring the vast space of chemical compounds and identifying novel molecules with optimal performance. However, in drug design problem, it is important to account for additional attributes, such as stability and toxicity, in addition to the target binding activity. In this talk, I will discuss the development and application of deep generative models for therapeutic discovery, which requires learning from data across multiple dimensions and generating novel molecules in a controllable manner.

<i>De novo</i> drug design with deep generative models

256th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.

De novo drug design with deep generative models