Timezone: »

 
Structure-based Drug Design with Equivariant Diffusion Models
Arne Schneuing · Yuanqi Du · Charles Harris · Arian Jamasb · Ilia Igashov · weitao Du · Tom Blundell · Pietro Lió · Carla Gomes · Max Welling · Michael Bronstein · Bruno Correia

@
in Machine Learning in Structural Biology Workshop »

Structure-based drug design (SBDD) aims to design small-molecule ligands that bind with high affinity and specificity to pre-determined protein targets. Traditional SBDD pipelines start with large-scale docking of compound libraries from public databases, thus limiting the exploration of chemical space to existent previously studied regions. Recent machine learning methods approached this problem using an atom-by-atom generation approach, which is computationally expensive. In this paper, we formulate SBDD as a 3D-conditional generation problem and present DiffSBDD, an E(3)-equivariant 3D-conditional diffusion model that generates novel ligands conditioned on protein pockets. Furthermore, we curate a new dataset of experimentally determined binding complex data from Binding MOAD to provide a realistic binding scenario that complements the synthetic CrossDocked dataset. Comprehensive in silico experiments demonstrate the efficiency of DiffSBDD in generating novel and diverse drug-like ligands that engage protein pockets with high binding energies as predicted by in silico docking.

[ Poster] 

Author Information

Arne Schneuing (École Polytechnique Fédérale de Lausanne)
Yuanqi Du (Cornell University)
Charles Harris (University of Cambridge)
Arian Jamasb (University of Cambridge)
Ilia Igashov (EPFL)
weitao Du (University of Science and Technology of China)
Tom Blundell (University of Cambridge)
Tom Blundell

Professor Sir Tom Blundell, FRS, FMedSci, is a Biochemistry Director of Research in Cambridge. He worked with Dorothy Hodgkin in Oxford in the 1960s on structure of insulin and then in Sussex on glucagon in the 1970s Recently he focusses on DNA repair, defining structures of multicomponent the 4100 amino acid DNA-PKcs complexes using cryo-EM. Over the past 30 years he has produced software for homology modelling, called Modeller cited 13,000 times, and for predicting impacts of mutations in cancer and drug resistance using AI/ML methods, contributing to ~700 research papers. In 1970s Tom developed structure-guided drug discovery, and in 1999 pioneered fragment-based drug discovery, co-founding Astex, with two oncology drugs on the market. In academia he develops antibiotics targeting mycobacteria in leprosy and cystic fibrosis. In 1970 as a councillor and Chair of Oxford City Planning, he stopped a motorway planned to go through the city centre and instead pedestrianized the area. He chaired UK Royal Commission on Environment, 1998 to 2005.

Pietro Lió (University of Cambridge)
Carla Gomes (Cornell University)
Max Welling (Microsoft Research AI4Science / University of Amsterdam)
Michael Bronstein (Oxford)
Bruno Correia (Ecole Polytechnique Federale de Lausanne)

More from the Same Authors