<ul><li>Protein binder design has seen advancements with hallucination-based methods optimizing structure prediction confidence metrics like interface predicted TM-score (ipTM) through backpropagation.</li><li>A new method has been proposed to extract statistical likelihoods from structure predictors by interpreting their confidence outputs as an energy-based model (EBM).</li><li>Leveraging the Joint Energy-based Modeling framework, pTMEnergy, a statistical energy function derived from predicted inter-residue error distributions, was introduced.</li><li>The incorporation of pTMEnergy into BindEnergyCraft (BECraft) design pipeline has shown superior performance over other methods, achieving higher in silico binder success rates and improved results in structure-based virtual screening tasks.</li></ul>

BindEnergyCraft: Casting Protein Structure Predictors as Energy-Based Models for Binder Design

Discover more