Biological Pathway Informed Models with Graph Attention Networks (GATs)

2025-08-30

AI Summary

Existing genomic models typically treat genes as unstructured labels or pathways merely as “gene sets,” neglecting their intrinsic topological structure and regulatory interactions. Method: We propose a pathway-aware modeling framework based on Graph Attention Networks (GAT), explicitly encoding intra-pathway regulatory relationships at the gene level using known biological pathway graphs; we further introduce an edge intervention mechanism to represent drug target perturbations, enabling dynamic rewiring of feedback loops. The method integrates hierarchical pathway modeling, training driven by temporal mRNA expression data, and mechanism-driven interpretability design. Contribution/Results: Experiments show an 81% reduction in MSE over an MLP baseline; the model successfully recapitulates all five known interactions in the TP53–MDM2–MDM4 pathway and significantly improves cross-condition generalizability and biological interpretability.

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: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Biological pathways map gene-gene interactions that govern all human processes. Despite their importance, most ML models treat genes as unstructured tokens, discarding known pathway structure. The latest pathway-informed models capture pathway-pathway interactions, but still treat each pathway as a "bag of genes" via MLPs, discarding its topology and gene-gene interactions. We propose a Graph Attention Network (GAT) framework that models pathways at the gene level. We show that GATs generalize much better than MLPs, achieving an 81% reduction in MSE when predicting pathway dynamics under unseen treatment conditions. We further validate the correctness of our biological prior by encoding drug mechanisms via edge interventions, boosting model robustness. Finally, we show that our GAT model is able to correctly rediscover all five gene-gene interactions in the canonical TP53-MDM2-MDM4 feedback loop from raw time-series mRNA data, demonstrating potential to generate novel biological hypotheses directly from experimental data.

Problem

• Modeling gene-level interactions within biological pathways
• Capturing pathway topology beyond bag-of-genes representations
• Predicting pathway dynamics under unseen treatment conditions

Innovation

• Graph Attention Networks model gene-level pathways
• Edge interventions encode drug mechanisms for robustness
• GATs capture gene-gene interactions from time-series data