What researchers are trying
IsoDDE reports a high-fidelity regime for difficult antibody–antigen interface prediction as part of the same all-atom engine used for small molecules.
High-fidelity antibody and biologic interfaces
Can general co-folding models accurately resolve antibody–antigen and other biologic interfaces with low sequence homology?
Why it matters
Accurate interfaces could guide epitope selection, affinity maturation, multispecific design and developability workflows before expensive structural experiments.
Evidence boundary
Benchmark composition, success thresholds and independent reproduction will determine how broadly the reported advantage generalizes.
Organizations represented
Isomorphic Labs
Demonstrated evidence
What has actually been shown.
- Developer-reported gains over AlphaFold 3 and Boltz-2 on a challenging low-homology test set.
Unresolved questions
- How does performance vary across CDR loops, nanobodies and multispecific formats?
- Can interface prediction improve prospective affinity maturation?
- How are glycosylation and conformational heterogeneity handled?
Signals to watch next
- Blind antibody benchmarks
- Prospective epitope design
- Experimental structures
- Developability integration
Connected evidence graph
Related BioAtlas model passports.
AlphaFold 2 / 3
The model that solved the 50-year protein-folding problem.
4/7 evidence fields documentedBoltz-1 / Boltz-2
Open-source AF3-quality structure — plus binding affinity.
4/7 evidence fields documentedChai-1 / Chai-2
An AlphaFold3-class complex predictor, made freely usable.
4/7 evidence fields documentedAbsci — Integrated Drug Creation
Generative de novo antibody design fused with wet-lab scale.
1/7 evidence fields documentedPrimary and evaluation sources