Mitigating Class Imbalance in Graph-Structured Data via Hierarchical Learning: Insights from Protein Binding Site Prediction
- 1Universidade Federal de Vicosa
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.
Related
- Hierarchical Temporal Transformer for Cancer Grade Prediction and Cross-Cancer Transfer Learning from Pathology Reports
- Hierarchical temporal transformer for cancer grade prediction and cross cancer transfer learning from pathology reports
- An extensible deep learning framework for multi-type chromatin interaction prediction through hierarchical adaptation strategies
- A Network-Structured Bayesian Hierarchical Model for Sparse Mutation-Drug Response Associations: Application to Cancer Pharmacogenomics
- Are You Learning Biological Signal or Shortcuts? Auditing and Mitigating Bias in Protein-Protein Interaction Datasets
BibTeX
@misc{paiva2026mitigating,
title = {{Mitigating Class Imbalance in Graph-Structured Data via Hierarchical Learning: Insights from Protein Binding Site Prediction}},
author = {V. Paiva and L. Marcolino and S. Izidoro and S. Silveira},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.28.754066},
url = {https://www.biorxiv.org/content/10.64898/2026.09.28.754066v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Mitigating Class Imbalance in Graph-Structured Data via Hierarchical Learning: Insights from Protein Binding Site Prediction
AU - V. Paiva
AU - L. Marcolino
AU - S. Izidoro
AU - S. Silveira
PY - 2026
DA - 2026/09/28
PB - bioRxiv
DO - 10.64898/2026.09.28.754066
UR - https://www.biorxiv.org/content/10.64898/2026.09.28.754066v1
AB - 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.
ER -