Conversion of a Viral Glycan Shield into a Binding Anchor via Causal-Driven Antibody Optimization

Conversion of a Viral Glycan Shield into a Binding Anchor via Causal-Driven Antibody Optimization The Antibody Evolution Nexus with Causal-Driven Simulation (AENCS) is a white-box computational-experimental framework for antibody engineering. By integrating physics-based simulation with causal inference, AENCS addresses a long-vacant frontier: the design of antibodies targeting glycosylated epitopes, where scarce and heterogeneous glycan structures render conventional approaches ineffective. Applied to the SARS-CoV-2 JN.1 variant, AENCS identified ACC01, an optimized antibody exhibiting ~24-fold improved neutralization. ACC01 stabilized the otherwise disordered N354 glycan, enabling its first cryo-EM structure determination, and functionally inverted this steric shield into a binding anchor through a multi-layered interaction network. This mechanistic understanding proved predictive, with ACC01 maintaining potent activity against the latest dominant variant NB.1.8.1. This work demonstrates that causal-driven antibody engineering can illuminate cryptic glycosylated epitopes, providing a viable paradigm for exploring this long-vacant frontier.