Mitigating Class Imbalance in Graph-Structured Data via Hierarchical Learning: Insights from Protein Binding Site Prediction

ICLR 2026 Conference Submission21756 Authors

Keywords: class imbalance, graph neural networks, hierarchical learning, subgraph classification, protein–ligand binding site prediction, structural bioinformatics

TL;DR: We propose a hierarchical graph learning framework that combines subgraph-level filtering with node-level classification, achieving improved performance on imbalanced graph benchmarks and protein–ligand binding site prediction.

Abstract: Learning from imbalanced data remains a major challenge for graph neural networks (GNNs), as minority nodes are not only rare but also structurally marginalized within the graph. We address this issue with CLARA, a hierarchical learning framework that decomposes node classification into two stages: a coarse subgraph-level classifier that selects regions likely to contain minority instances, followed by a fine-grained node-level predictor within these regions. This design improves sensitivity while maintaining scalability, filtering out irrelevant areas and focusing learning on topologically meaningful neighborhoods. Experiments on benchmark graph datasets demonstrate substantial gains over established imbalance-handling methods, with CLARA reaching an F1-score of 88.3%. The same strategy achieves significant improvements in protein–ligand binding site prediction, underscoring its broad and consistent effectiveness across both biological and general graph learning tasks.

Primary Area: learning on graphs and other geometries & topologies

Submission Number: 21756