Evolving Robust Drug Candidates via Co-Evolutionary Artificial Life Simulators

David Scott Lewis, Enrique Zueco, Junhan Wang

Submission Type: I want my submission to be considered for both oral and poster presentation.

Keywords: drug discovery, AI for chemistry, AI for biology, co-evolutionary algorithms, resistance-aware design, artificial life, Red Queen dynamics, graph neural networks, structure-based drug design, robust drug scaffolds, mutating targets, escape mutation, NK fitness landscapes, surrogate fitness oracles, evolutionary search, autonomous chemistry, self-driving labs, in silico screening, resistance prediction, molecular graphs, adaptive therapeutics, closed-loop discovery, benchmark simulation, autonomous agents for science

TL;DR: Drug design is recast as Red Queen-style co-evolution: evolving compounds and mutating protein targets adapt against each other in simulation, allowing a fast GNN fitness model to identify scaffolds that remain effective as resistance emerges.

Confirmation Of Submission Requirements: I submit an abstract. It uses the template provided on the submission page and is no longer than 2 pages.

Submission Number: 404