MOTIVATION: Predicting immunoglobulin-antigen (Ig-Ag) binding remains a significant challenge due to the paucity of experimentally resolved complexes and the limited accuracy of de novo Ig structure prediction. RESULTS: We introduce IgPose, a generalizable framework for Ig-Ag pose identification and scoring, built on a generative data-augmentation pipeline. To mitigate data scarcity, we constructed the Structural Immunoglobulin Decoy Database (SIDD), a comprehensive repository of high-fidelity synthetic decoys. IgPose integrates equivariant graph neural networks, ESM-2 embeddings, and gated recurrent units to synergistically capture both geometric and evolutionary features. We implemented interface-focused k-hop sampling with biologically guided pooling to enhance generalization across diverse interfaces. The framework comprises two sub-networks-IgPoseClassifier for binding pose discrimination and IgPoseScore for DockQ score estimation-and achieves robust performance on curated internal test sets and the CASP-16 benchmark compared to physics and deep learning baselines. IgPose serves as a versatile computational tool for high-throughput antibody discovery pipelines by providing accurate pose filtering and ranking. AVAILABILITY AND IMPLEMENTATION: IgPose is available on GitHub (https://github.com/arontier/igpose).

Graph Neural Networks have achieved notable success, yet explaining their rationales remains a challenging problem. Existing methods, including post-hoc and interpretable approaches, have numerous limitations. Post-hoc methods treat models as black boxes and can mislead users, while interpretable models often overlook user-centric explanations. Furthermore, most existing methods do not carefully consider the user’s perception of explanations, potentially resulting in explanation-user mismatches. To address these problems, we propose a novel interpretable concept-matching model to enhance GNN interpretability and prediction accuracy. The proposed model extracts frequent concepts from input graphs using the graph information bottleneck theory and modified constraints. These concepts are managed in an in-memory concept corpus for efficient inference lookups and explanation generation. Various explanation construction features are implemented based on the concept corpus and the discovery module, aiming to fulfill diverse user preferences. Extensive experiments and a user study validate the performance of the proposed approach, showcasing its potential for improving model accuracy and interpretability.