MERGE: Multi-faceted Hierarchical Graph-based GNN for Gene Expression Prediction from Whole Slide Histopathology Images

2024-12-03

arXiv.org

AI Summary

Existing methods struggle to model spatial and morphological interactions across regions in whole-slide histopathology images (WSIs), limiting the accuracy of gene expression prediction. To address this, we propose a multi-dimensional hierarchical graph modeling framework: (1) tissue regions are first partitioned via multi-scale image clustering; (2) a spatial-morphological joint graph is constructed, integrating intra- and inter-cluster connections to explicitly encode both local and long-range tissue associations; (3) a hierarchical graph neural network (GNN) is designed to enable cross-region collaborative expression inference; and (4) a gene-aware data smoothing strategy is introduced to enhance biological plausibility. This work establishes the first end-to-end graph learning paradigm for WSI-to-gene-expression mapping grounded in spatial transcriptomics. Evaluated on multiple benchmark datasets, our method achieves significant improvements over state-of-the-art approaches in MSE, Pearson correlation, and robustness.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

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

Abstract

Recent advances in Spatial Transcriptomics (ST) pair histology images with spatially resolved gene expression profiles, enabling predictions of gene expression across different tissue locations based on image patches. This opens up new possibilities for enhancing whole slide image (WSI) prediction tasks with localized gene expression. However, existing methods fail to fully leverage the interactions between different tissue locations, which are crucial for accurate joint prediction. To address this, we introduce MERGE (Multi-faceted hiErarchical gRaph for Gene Expressions), which combines a multi-faceted hierarchical graph construction strategy with graph neural networks (GNN) to improve gene expression predictions from WSIs. By clustering tissue image patches based on both spatial and morphological features, and incorporating intra- and inter-cluster edges, our approach fosters interactions between distant tissue locations during GNN learning. As an additional contribution, we evaluate different data smoothing techniques that are necessary to mitigate artifacts in ST data, often caused by technical imperfections. We advocate for adopting gene-aware smoothing methods that are more biologically justified. Experimental results on gene expression prediction show that our GNN method outperforms state-of-the-art techniques across multiple metrics.

Problem

• Predict gene expression from histopathology images
• Leverage tissue location interactions for accurate predictions
• Mitigate artifacts in Spatial Transcriptomics data

Innovation

• Multi-faceted hierarchical graph construction strategy
• Graph neural networks for gene expression prediction
• Gene-aware smoothing to mitigate ST data artifacts