Antibody-Antigen Affinity Prediction with Chain-Aware Protein Language Modeling
- 1Shiv Nadar Institution of Eminence (Deemed to be University), Delhi-NCR
Abstract
Motivation: Antibody-antigen affinity determines which antibodies advance in therapeutic discovery, repertoire analysis and affinity maturation, but experimental measurements are sparse relative to the scale of sequence libraries. Structure-based predictors can exploit interface geometry when reliable complexes are available, yet early discovery often requires ranking many heavy-light chain pairs against antigens for which no complex structure exists. Existing sequence-based models are scalable, but frequently compress heavy and light chains into a single antibody representation or concatenate antibody and antigen features, potentially obscuring the chain-specific and epitope-specific signals that drive binding. Results: We present AbAffinity, a sequence-only chain-aware three-stream architecture that maintains heavy chain, light chain and antigen as distinct streams. It integrates frozen ESM-2 embeddings with heavy-chain CDR-focused pooling, heavy-light self-attention, adaptive fusion and gated cross-attention, training only a compact interaction module. Under nested ten-fold cross-validation on SAAINT-DB, AbAffinity achieves Pearson r = 0.838 {+/-} 0.033, Spearman {rho} = 0.828 {+/-} 0.027 and RMSE = 0.738 {+/-} 0.078. AbAffinity maintains meaningful performance on stricter PDB-disjoint and antigen sequence-identity splits, and held-out complexes. Evaluation on external natural-complex benchmarks further supports consistent transfer. Partner-shuffling, attribution and uncertainty analyses support partner-conditioned prediction and biologically plausible residue-level hypotheses. AbAffinity provides a lightweight, explainable sequence-first framework for antibody triage when structural information is limited or unavailable.
Related
- CIR-DDG: backbone-agnostic residual correction of antibody-antigen affinity changes with explicit cross-chain geometry
- Benchmarking Large Language Models for Assay-Aware Drug-Target Binding Affinity Prediction
- Sequence-only antigen context improves modeling of antibody hypervariable regions
- Decoding antibody-antigen recognition: the BL8 language and its application in calibrating deep learning predictions
- BCell-Epitope-Prediction: A local, reproducible pipeline for multi-epitope vaccine design: integration and structural validation against a panel of PDB antigen-antibody complexes
BibTeX
@misc{singh2026antibodyantigen,
title = {{Antibody-Antigen Affinity Prediction with Chain-Aware Protein Language Modeling}},
author = {H. Singh and A. Malhotra and S. P. Srivastava and R. K. SINGH and R. Gorantla},
howpublished = {bioRxiv},
year = {2026},
month = jun,
doi = {10.64898/2026.06.19.733375},
url = {https://www.biorxiv.org/content/10.64898/2026.06.19.733375v2},
note = {CC BY},
}
RIS
TY - UNPB
TI - Antibody-Antigen Affinity Prediction with Chain-Aware Protein Language Modeling
AU - H. Singh
AU - A. Malhotra
AU - S. P. Srivastava
AU - R. K. SINGH
AU - R. Gorantla
PY - 2026
DA - 2026/06/19
PB - bioRxiv
DO - 10.64898/2026.06.19.733375
UR - https://www.biorxiv.org/content/10.64898/2026.06.19.733375v2
AB - Motivation: Antibody-antigen affinity determines which antibodies advance in therapeutic discovery, repertoire analysis and affinity maturation, but experimental measurements are sparse relative to the scale of sequence libraries. Structure-based predictors can exploit interface geometry when reliable complexes are available, yet early discovery often requires ranking many heavy-light chain pairs against antigens for which no complex structure exists. Existing sequence-based models are scalable, but frequently compress heavy and light chains into a single antibody representation or concatenate antibody and antigen features, potentially obscuring the chain-specific and epitope-specific signals that drive binding. Results: We present AbAffinity, a sequence-only chain-aware three-stream architecture that maintains heavy chain, light chain and antigen as distinct streams. It integrates frozen ESM-2 embeddings with heavy-chain CDR-focused pooling, heavy-light self-attention, adaptive fusion and gated cross-attention, training only a compact interaction module. Under nested ten-fold cross-validation on SAAINT-DB, AbAffinity achieves Pearson r = 0.838 {+/-} 0.033, Spearman {rho} = 0.828 {+/-} 0.027 and RMSE = 0.738 {+/-} 0.078. AbAffinity maintains meaningful performance on stricter PDB-disjoint and antigen sequence-identity splits, and held-out complexes. Evaluation on external natural-complex benchmarks further supports consistent transfer. Partner-shuffling, attribution and uncertainty analyses support partner-conditioned prediction and biologically plausible residue-level hypotheses. AbAffinity provides a lightweight, explainable sequence-first framework for antibody triage when structural information is limited or unavailable.
ER -