University of Cambridge > Talks.cam > Artificial Intelligence Research Group Talks (Computer Laboratory) > Structure-aware generation of molecules in protein pockets

Structure-aware generation of molecules in protein pockets

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Mateja Jamnik.

Join us on Zoom

Searching the vast chemical space for potential drugs is intractable even for computational screening approaches, though deep generative methods have shown promise in proposing novel molecules from scratch (de-novo design). However, most models fail to incorporate 3D protein structures that are salient in designing drugs that bind to protein targets. I present a novel supervised approach that combines insights from graph-generative and 3D models and constructs drugs inside protein pockets, atom-by-atom, guided by structural information from crystallographic drug datasets. I find that the model improves binding affinities on average by 8% and druglikeness scores by 10% over the non-guided baseline, and proposes some molecules that exceed the binding scores of known ligands.

This talk is part of the Artificial Intelligence Research Group Talks (Computer Laboratory) series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2021 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity