Molecular design improves by considering many 3d shapes at once
Ensemble-Conditioned Molecular Design
Machine LearningNeural and Evolutionary Computing
Summary
Designing molecules often focuses on finding just one shape that fits a target, but in reality, molecules wiggle into many shapes. The authors propose a new way to design molecules by looking at the whole range of shapes and their properties together. They built a method that lets a computer create molecules that meet multiple shape and property goals simultaneously. Their approach works better for tricky tasks, like making molecules that can bind to two targets or activate only certain protein states.
What this means in practice
- •For drug designers: Create molecules optimized to bind multiple targets or protein states by conditioning on several 3D shape and property requirements.
- •For biotech engineers: Generate flexible-size molecular structures meeting complex 3D constraints for experimental synthesis and testing.
Authors
Ross Irwin, Alessandro Tibo, Jon Paul Janet, Simon Olsson
Abstract
Molecular design is typically approached as a problem of finding molecules which can adopt a single bioactive conformation. In reality, molecules occupy a distribution over conformations, and many of the properties which determine whether a candidate is viable depend on that distribution rather than on any single conformer. We reframe molecular design as an optimisation of both the modes and properties of molecules' conformational ensembles, where modes can be represented as shapes, pharmacophore profiles or protein pockets, and properties are aggregate scalars computed over the whole distribution. To realise this we introduce ensemble-conditioned guidance, a framework which conditions 3D molecular generative models on both axes simultaneously. Mode conditions are composed adaptively at inference by combining the vector fields produced under each condition. Conditions may be targeted or avoided, mixed across modalities and combined in arbitrary numbers, allowing a wide range of design tasks to be expressed with a single trained model. We introduce adaptive symmetry learning to allow conditions from different reference frames to be composed, and extend our generative framework to enable flexible-size generation. We evaluate on new benchmarks for multi-mode conditioning and ensemble property optimisation, and apply the framework to two practical drug discovery tasks, dual-target binder design and active-state-selective agonist design, where in both cases conditioning on the additional state improves the desired outcome over single-state conditioning.