Predicting Epitopes of Synthetic Binding Proteins Using Protein Structure Prediction and Protein-Protein Docking

Synthetic binding proteins (SBPs) engineered from privileged protein scaffolds usually have high specificity to protein target. However, epitope information of currently known SBPs is still limited, which hinders the development of protein binders with desired function. Herein, a framework combining protein structure prediction (AlphaFold2) and protein-protein docking (HawkDock and RosettaDock) was designed for predicting the epitope of specific SBP. Using the experimental structure of a nanobody in complex with SARS-CoV-2 receptor binding domain (RBD) as an example, the computational framework perfectly reproduced the binding mode between the SBP and target protein. Therefore, the strategy can be expanded to predict the epitopes of other SBPs, which may provide useful information for understanding the molecular mechanism of protein-protein recognition, but also facilitate the rational design of novel SBPs with enhanced properties.