GNN-Suite: a Graph Neural Network Benchmarking Framework for Biomedical Informatics

2025-05-15

AI Summary

Standardized and reproducible evaluation of graph neural networks (GNNs) in biomedical informatics—particularly for cancer driver gene identification—remains lacking. Method: This paper introduces the first modular GNN benchmark framework tailored to driver gene prediction. It (1) adopts a Nextflow-based workflow to ensure fully reproducible experimental evaluation; (2) proposes a unified graph construction and multimodal feature fusion strategy integrating heterogeneous molecular interaction networks (STRING, BioGRID, PID, COSMIC); and (3) establishes a statistically robust benchmark protocol encompassing multiple GNN architectures (GCN2, GAT, GIN, GraphSAGE), 10 independent random seed runs, and rigorous performance aggregation. Results: GCN2 achieves the highest balanced accuracy (0.807 ± 0.035) on the STRING-derived graph; all GNNs significantly outperform logistic regression baselines, demonstrating that explicit modeling of network topology yields substantial and consistent gains in driver gene identification.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

We present GNN-Suite, a robust modular framework for constructing and benchmarking Graph Neural Network (GNN) architectures in computational biology. GNN-Suite standardises experimentation and reproducibility using the Nextflow workflow to evaluate GNN performance. We demonstrate its utility in identifying cancer-driver genes by constructing molecular networks from protein-protein interaction (PPI) data from STRING and BioGRID and annotating nodes with features from the PCAWG, PID, and COSMIC-CGC repositories. Our design enables fair comparisons among diverse GNN architectures including GAT, GAT3H, GCN, GCN2, GIN, GTN, HGCN, PHGCN, and GraphSAGE and a baseline Logistic Regression (LR) model. All GNNs were configured as standardised two-layer models and trained with uniform hyperparameters (dropout = 0.2; Adam optimiser with learning rate = 0.01; and an adjusted binary cross-entropy loss to address class imbalance) over an 80/20 train-test split for 300 epochs. Each model was evaluated over 10 independent runs with different random seeds to yield statistically robust performance metrics, with balanced accuracy (BACC) as the primary measure. Notably, GCN2 achieved the highest BACC (0.807 +/- 0.035) on a STRING-based network, although all GNN types outperformed the LR baseline, highlighting the advantage of network-based learning over feature-only approaches. Our results show that a common framework for implementing and evaluating GNN architectures aids in identifying not only the best model but also the most effective means of incorporating complementary data. By making GNN-Suite publicly available, we aim to foster reproducible research and promote improved benchmarking standards in computational biology. Future work will explore additional omics datasets and further refine network architectures to enhance predictive accuracy and interpretability in biomedical applications.

Problem

• Standardizing GNN benchmarking for biomedical informatics
• Identifying cancer-driver genes using molecular networks
• Enabling fair comparisons among diverse GNN architectures

Innovation

• Modular framework for GNN benchmarking
• Standardized workflow with Nextflow
• Uniform hyperparameters for fair comparison