Fast, accurate ranking of engineered proteins by receptor binding propensity using structural modeling

Deep learning-based methods for protein structure prediction have achieved unprecedented accuracy. However, the power of these tools to guide the engineering of protein-based therapeutics remains limited due to a gap between the ability to predict the structures of candidate proteins and the ability to assess which of those proteins are most likely to bind to a target receptor. Here we bridge this gap by introducing Automated Pairwise Peptide-Receptor AnalysIs for Screening Engineered proteins (APPRAISE), a method for predicting the receptor binding propensity of engineered proteins. After generating models of engineered proteins competing for binding to a target using an established structure-prediction tool such as AlphaFold2-multimer or ESMFold, APPRAISE performs a rapid (under 1 CPU second per model) scoring analysis that takes into account biophysical and geometrical constraints. As a proof-of-concept, we demonstrate that APPRAISE can accurately classify receptor-dependent vs. receptor-independent engineered adeno-associated viral vectors, as well as diverse classes of engineered proteins such as miniproteins targeting the SARS-CoV-2 spike protein, nanobodies targeting a G-protein-coupled receptor, and peptides that specifically bind to transferrin receptor and PD-L1. With its high accuracy, interpretability, and generalizability, APPRAISE has the potential to expand the utility of current structural prediction and accelerate protein engineering for biomedical applications.