Objectives: Build docking, shape-phonon and ML models that integrate WP2 cryo-EM/smFRET states with WP1 assay outcomes; train classifiers to penalise ATP-like geometries and prioritise interface-compatible scaffolds; generate prospective design hypotheses and SAR expansions for DC6–DC8 discovery streams.
Expected results: Open, validated modelling pipelines and benchmark datasets; a methods paper demonstrating improved enrichment for shallow kinase interfaces; a prospective case study showing hit-rate uplift in WP3 screens.
Secondments: EPFL: Experimental grounding in peptide interface binders (M20–23; 4 months); KUL: Testing DC9-prioritised designs in the bisubstrate assay (M31–33; 3 months).
Integration: Provides the central computational design loop linking WP2 structural signatures, WP1 assay labels and WP3 screening/optimisation workflows.
Interdisciplinarity: Combines cheminformatics & machine learning, phage-display screening and assay-bases hit triage.
