Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
AbGaze addresses the failure of scalar distance-based representations in antibody design by explicitly modeling the distance, spatial direction, and surface-normal orientation of antigen surfaces within antibody-residue local frames. By coupling this attentive geometric representation with local-frame supervision, the framework jointly tackles multi-CDR co-design, complex structure prediction, and affinity optimization. The approach reduces CDR structural error by 14.9%, boosts sequence recovery by 7.1%, and yields a 32.5% increase in binding affinity improvement rate over state-of-the-art baselines.
Explicit cross-molecular surface orientations—rather than standard pairwise scalar distances—unlock a 32.5% surge in antibody affinity optimization while cutting CDR structural errors by 15%.
Computational antibody design requires representations that capture the geometric patterns underlying antigen--antibody interactions, yet existing approaches often rely on scalar distances or surface-intrinsic features, leaving cross-molecular geometry largely implicit. We present AbGaze, an end-to-end antibody design framework based on attentive geometric representation learning, which encodes distance, spatial direction, and surface-normal orientation of antigen surfaces relative to antibody-residue local frames, and adaptively aggregates these geometric interactions according to their interfacial context. The learned interaction representation is shared across multi-CDR co-design, complex structure prediction, and affinity optimization, with local-frame geometric supervision further constraining the representation. AbGaze outperforms prior methods across all three tasks: relative to the second-best method, it improves amino-acid recovery by 7.1% and reduces structural error by 14.9% on average over the six CDRs, improves interface docking quality (DockQ) by 6.6%, and raises the affinity improvement rate (IMP) by 32.5%.
Tsinghua AI