Fast, accurate antibody structure prediction from deep learning on massive set of natural antibodies

Abstract

Antibodies have the capacity to bind a diverse set of antigens and have become critical therapeutics and diagnostic molecules. The binding of antibodies is facilitated by a set of six hypervariable loops that are diversified through genetic recombination and mutation. Accurate structural modeling of these loops is critical for rational design of antibodies, but remains an expensive, time-consuming endeavor using traditional experimental methods. Even with recent advances, accurate computational prediction of these hypervariable loops remains a challenge.