CIR-DDG: backbone-agnostic residual correction of antibody-antigen affinity changes with explicit cross-chain geometry
Abstract
Motivation: Accurate prediction of mutation-induced protein--protein binding free-energy changes is important for antibody affinity maturation, yet scarce labels and complex interface geometry limit generalization. Heterogeneous predictors may process three-dimensional complexes without preserving the cross-chain signals most relevant to a mutation in their final scalar output. Results: We introduce CIR-DDG, a lightweight residual adapter that combines a fixed base prediction with 22 interpretable descriptors of cross-chain distance, contact density and site--partner context. In complex-level five-fold evaluation on SKEMPI 2.0 measurements from 343 complexes, CIR-DDG improved all six tested backbones on antibody--antigen interface mutations: Spearman correlation increased by 0.0346--0.1296, while RMSE decreased by 0.0074--0.0408\kcalmol. Cross-validated probing, equal-capacity controls and feature ablations support the complementarity of explicit geometry. On an independent SARS-CoV-2 RBD--ACE2 deep-mutational-scan benchmark of 3669 substitutions, the fold-specific adapters transferred without any retraining: the absolute interface Spearman correlation increased by 0.026--0.081 for all four evaluable backbones, showing that the learned geometric correction generalizes beyond SKEMPI thermodynamic measurements. Availability and implementation: CIR-DDG is available at https://github.com/ecnuabmlab/CIR-ddG.
Related
- Optimal Maturation Protocols for High-Affinity Antibody Targets: A Path-Integral Approach
- Antibody-Antigen Affinity Prediction with Chain-Aware Protein Language Modeling
- Active Learning framework for Antibody Binding Affinity Prediction
- Automatic Quality Control and Error Correction in MRI linear registration via a Residual Parameter Prediction Network for T1w MRI
- Panic at the Discordance: the Asparagaceae Phylogenetic Backbone
BibTeX
@misc{yu2026cirddg,
title = {{CIR-DDG: backbone-agnostic residual correction of antibody-antigen affinity changes with explicit cross-chain geometry}},
author = {Weilun Yu and Zhiheng Zou and Yonggui Huang and Honggang Qi and Geguang Pu and Yu Xiao and Gang Xu and Jiangtao Wang and Xi Chen},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2608.27530},
eprint = {2608.27530},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.27530},
}
RIS
TY - UNPB
TI - CIR-DDG: backbone-agnostic residual correction of antibody-antigen affinity changes with explicit cross-chain geometry
AU - Weilun Yu
AU - Zhiheng Zou
AU - Yonggui Huang
AU - Honggang Qi
AU - Geguang Pu
AU - Yu Xiao
AU - Gang Xu
AU - Jiangtao Wang
AU - Xi Chen
PY - 2026
DA - 2026/08/27
PB - arXiv
DO - 10.48550/arxiv.2608.27530
UR - https://arxiv.org/abs/2608.27530
AB - Motivation: Accurate prediction of mutation-induced protein--protein binding free-energy changes is important for antibody affinity maturation, yet scarce labels and complex interface geometry limit generalization. Heterogeneous predictors may process three-dimensional complexes without preserving the cross-chain signals most relevant to a mutation in their final scalar output. Results: We introduce CIR-DDG, a lightweight residual adapter that combines a fixed base prediction with 22 interpretable descriptors of cross-chain distance, contact density and site--partner context. In complex-level five-fold evaluation on SKEMPI 2.0 measurements from 343 complexes, CIR-DDG improved all six tested backbones on antibody--antigen interface mutations: Spearman correlation increased by 0.0346--0.1296, while RMSE decreased by 0.0074--0.0408\kcalmol. Cross-validated probing, equal-capacity controls and feature ablations support the complementarity of explicit geometry. On an independent SARS-CoV-2 RBD--ACE2 deep-mutational-scan benchmark of 3669 substitutions, the fold-specific adapters transferred without any retraining: the absolute interface Spearman correlation increased by 0.026--0.081 for all four evaluable backbones, showing that the learned geometric correction generalizes beyond SKEMPI thermodynamic measurements. Availability and implementation: CIR-DDG is available at https://github.com/ecnuabmlab/CIR-ddG.
ER -