PEaRL: Pathway-Enhanced Representation Learning for Gene and Pathway Expression Prediction from Histology

2025-10-03

AI Summary

This study addresses two key bottlenecks in histology-based gene and pathway expression prediction: (1) existing methods focus predominantly on highly variable genes, neglecting biologically coherent functional units; and (2) they ignore coordinated regulation at the pathway level. To this end, we propose the first pathway-oriented multimodal modeling framework: it represents transcriptomic profiles via ssGSEA-derived pathway activation scores, encodes histological features using a Transformer architecture, and employs contrastive learning to align image and pathway-space representations. Crucially, our approach explicitly embeds functional pathway signals into cross-modal modeling, substantially enhancing interpretability and biological plausibility. Evaluated on three cancer spatial transcriptomics datasets, our method achieves up to 58.9% and 20.4% improvements in Spearman correlation over state-of-the-art methods for gene-level and pathway-level expression prediction, respectively. Moreover, it enables effective dimensionality reduction while ensuring cross-modal semantic consistency.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Integrating histopathology with spatial transcriptomics (ST) provides a powerful opportunity to link tissue morphology with molecular function. Yet most existing multimodal approaches rely on a small set of highly variable genes, which limits predictive scope and overlooks the coordinated biological programs that shape tissue phenotypes. We present PEaRL (Pathway Enhanced Representation Learning), a multimodal framework that represents transcriptomics through pathway activation scores computed with ssGSEA. By encoding biologically coherent pathway signals with a transformer and aligning them with histology features via contrastive learning, PEaRL reduces dimensionality, improves interpretability, and strengthens cross-modal correspondence. Across three cancer ST datasets (breast, skin, and lymph node), PEaRL consistently outperforms SOTA methods, yielding higher accuracy for both gene- and pathway-level expression prediction (up to 58.9 percent and 20.4 percent increase in Pearson correlation coefficient compared to SOTA). These results demonstrate that grounding transcriptomic representation in pathways produces more biologically faithful and interpretable multimodal models, advancing computational pathology beyond gene-level embeddings.

Problem

• Predicts gene and pathway expression from histology images
• Overcomes limitations of small gene sets in multimodal approaches
• Improves interpretability and accuracy in computational pathology

Innovation

• Uses pathway activation scores for transcriptomics representation
• Aligns pathway signals with histology via contrastive learning
• Encodes biological pathways using transformer architecture