My research focuses on developing structure-based computational strategies for antibody design. Computational approaches can overcome the time-consuming, costly, and imprecise nature of traditional experimental techniques, which often yield antibodies with suboptimal functionality and developability. Past methods fall short of matching the affinity and specificity of clinically-approved antibodies. My work aims to develop machine learning methods that design epitope-specific antibodies with improved binding properties.
Publications
Beyond AlphaFold 3: a standardised benchmark of open-source co-folding models for antibody-antigen complexes, from docking accuracy to interaction specificity.
MAbs
(2026)
18
2760737
Deep learning assessment of nativeness and pairing likelihood for antibody and nanobody design with AbNatiV2.
Mabs
(2026)
18
2646361
Disulphide and sequence-encoded conformational priors guide nanobody structure prediction.
(2026)
(doi: 10.64898/2026.02.13.705647)
Improving nanobody structure prediction with self-distillation
(2025)
(doi: 10.64898/2025.12.01.691162)
Deep learning assessment of nativeness and pairing likelihood for antibody and nanobody design with AbNatiV2
(2025)
(doi: 10.1101/2025.10.31.685806)