Skip to yearly menu bar Skip to main content


Poster
in
Workshop: AI for Science: Progress and Promises

A 3D-Shape Similarity-based Contrastive Approach to Molecular Representation Learning

Austin Atsango · Nathaniel Diamant · Ziqing Lu · Tommaso Biancalani · Gabriele Scalia · Kangway Chuang

Keywords: [ Representation Learning ] [ graph neural networks ] [ shape similarity ] [ conformers ] [ three-dimensional representations ]


Abstract:

Molecular shape and geometry dictate key biophysical recognition processes, yet many modern graph neural networks disregard 3D information for molecular property prediction. Here, we propose a new contrastive-learning procedure for graph neural networks, Molecular Contrastive Learning from Shape Similarity (MolCLaSS), that implicitly learns a three-dimensional representation. Rather than directly encoding or targeting three-dimensional poses, MolCLaSS matches a similarity objective based on Gaussian overlays to learn a meaningful representation of molecular shape. We demonstrate how this framework naturally captures key aspects of three-dimensionality that two-dimensional representations cannot and provides an inductive framework for scaffold hopping.

Chat is not available.