Supervised Graph Contrastive Learning for Gene Regulatory Network

2025-05-23

AI Summary

Existing graph contrastive learning (GCL) methods applied to gene regulatory networks (GRNs) neglect biologically realistic perturbations—such as gene knockdowns—leading to representations lacking interpretability and biological relevance. To address this, we propose SupGCL, the first supervised GCL framework that explicitly incorporates gene knockdown experiments as supervisory signals into contrastive learning, yielding an interpretable, probabilistic contrastive model. SupGCL integrates perturbation-driven data augmentation, probabilistic graph modeling, and multi-task downstream adaptation—supporting both graph-level and node-level tasks. Evaluated on GRNs derived from multiple cancer patient cohorts, SupGCL consistently outperforms state-of-the-art methods across patient risk prediction, disease subtype classification, and gene functional annotation. Our approach establishes a novel paradigm for biologically grounded representation learning on regulatory networks.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Graph representation learning is effective for obtaining a meaningful latent space utilizing the structure of graph data and is widely applied, including biological networks. In particular, Graph Contrastive Learning (GCL) has emerged as a powerful self-supervised method that relies on applying perturbations to graphs for data augmentation. However, when applying existing GCL methods to biological networks such as Gene Regulatory Networks (GRNs), they overlooked meaningful biologically relevant perturbations, e.g., gene knockdowns. In this study, we introduce SupGCL (Supervised Graph Contrastive Learning), a novel GCL method for GRNs that directly incorporates biological perturbations derived from gene knockdown experiments as the supervision. SupGCL mathematically extends existing GCL methods that utilize non-biological perturbations to probabilistic models that introduce actual biological gene perturbation utilizing gene knockdown data. Using the GRN representation obtained by our proposed method, our aim is to improve the performance of biological downstream tasks such as patient hazard prediction and disease subtype classification (graph-level task), and gene function classification (node-level task). We applied SupGCL on real GRN datasets derived from patients with multiple types of cancer, and in all experiments SupGCL achieves better performance than state-of-the-art baselines.

Problem

• Incorporates gene knockdown data into graph contrastive learning
• Improves biological network representation for downstream tasks
• Enhances performance in cancer patient hazard prediction

Innovation

• Incorporates gene knockdown data as supervision
• Extends GCL with probabilistic biological models
• Improves performance in biological downstream tasks