AbBiBench (Antibody Binding Benchmark)
Benchmark for antibody binding affinity maturation treating Ab-Ag complex as fundamental unit. 184,500+ experimental measurements across 9 antigen targets (influenza, SARS-CoV-2, HER2, VEGF). Compares language models, inverse folding, diffusion, and geometric models.
Composite
80.8
Experimental validation
Retrospective
Stages
Lead ID / ADMET
Modalities
protein_structureantibody
Task types
binding_affinityaffinity_maturationprotein_design
Size
entries: 184
measurements: 500
antigens: 9
measurements: 500
antigens: 9
License
Unknown
First release
2025-05
Last updated
2025-10
Official site
→ project page
Leaderboard
→ leaderboard
Dataset
→ dataset
Code / GitHub
→ repository
HuggingFace
→ HF
Paper
AbBiBench: A Benchmark for Antibody Binding Affinity Maturation and Design · · 2025 · paper · 4 citations
Flags
biologicsaffinity_maturation
Experts
—
Groups
—
Hosted by
—
Related benchmarks
Rubric (7-criterion)
rigor
5
coverage
4
maintenance
3
adoption
3
quality
4
accessibility
4
industry_relevance
5
Notes
Key finding: structure-conditioned inverse folding models outperform others for affinity prediction and generation. Treats complex holistically rather than antibody in isolation. Demonstrates utility in generative design for H1N1 antibody improvement.