Hypergraph Neural Networks Reveal Spatial Domains from Single-cell Transcriptomics Data

2024-10-23

arXiv.org

AI Summary

Traditional graph neural networks (GNNs) applied to single-cell spatial transcriptomics data model only pairwise relationships, failing to capture higher-order cooperative cellular affiliations and thus yielding inaccurate spatial domain identification. To address this, we propose the first unsupervised hypergraph autoencoder framework for spatial domain discovery. Our method constructs a hypergraph by jointly encoding spatial neighborhood and gene expression similarity, explicitly modeling high-order cellular co-localization. We innovatively integrate hypergraph neural networks (HGNNs) with variational autoencoders (VAEs) to enable end-to-end representation learning and clustering optimization. On standard benchmarks, our approach achieves substantial improvements: iLISI = 1.843 (+12.7%), ARI = 0.51, and Leiden score = 0.60. Moreover, it enhances biological coherence of identified spatial domains and preserves greater cellular type diversity.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• 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: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

The task of spatial clustering of transcriptomics data is of paramount importance. It enables the classification of tissue samples into diverse subpopulations of cells, which, in turn, facilitates the analysis of the biological functions of clusters, tissue reconstruction, and cell-cell interactions. Many approaches leverage gene expressions, spatial locations, and histological images to detect spatial domains; however, Graph Neural Networks (GNNs) as state of the art models suffer from a limitation in the assumption of pairwise connections between nodes. In the case of domain detection in spatial transcriptomics, some cells are found to be not directly related. Still, they are grouped as the same domain, which shows the incapability of GNNs for capturing implicit connections among the cells. While graph edges connect only two nodes, hyperedges connect an arbitrary number of nodes along their edges, which lets Hypergraph Neural Networks (HGNNs) capture and utilize richer and more complex structural information than traditional GNNs. We use autoencoders to address the limitation of not having the actual labels, which are well-suited for unsupervised learning. Our model has demonstrated exceptional performance, achieving the highest iLISI score of 1.843 compared to other methods. This score indicates the greatest diversity of cell types identified by our method. Furthermore, our model outperforms other methods in downstream clustering, achieving the highest ARI values of 0.51 and Leiden score of 0.60.

Problem

• Detect spatial domains in single-cell transcriptomics data
• Overcome GNN limitations in capturing implicit cell connections
• Improve clustering accuracy and diversity in unsupervised learning

Innovation

• Hypergraph Neural Networks capture complex cell connections
• Autoencoders enable unsupervised learning without labels
• Model achieves highest diversity and clustering scores