# MMSpa is a deep learning-based tool that enhances the identification of spatial domains in spatial transcriptomics studies

## TL;DR

MMSpa is a new deep learning tool that improves the identification of spatial domains in spatial transcriptomics by better capturing gene expression patterns and tissue boundaries.

## Contribution

MMSpa introduces a masked graph attention autoencoder with an edge-removal strategy to enhance spatial domain identification in transcriptomics.

## Key findings

- MMSpa outperforms existing methods in identifying spatial domains across multiple ST technologies and platforms.
- MMSpa excels in detecting finer-grained functional tissue domains in complex and heterogeneous tissues.
- The method reveals clearer domain boundaries and compensates for the lack of spatial annotation in histopathology.

## Abstract

Spatial transcriptome (ST) technologies have transformed the study of tissue structure by retaining the spatial distribution of gene expression. One major challenge in accurately identifying spatial domains is to extract domain-related information from spatial locations and gene expression. Here, we propose MMSpa, a masked graph attention autoencoder framework specifically designed to improve spatial domain identification. MMSpa incorporates an edge-removal strategy to construct an enhanced spatial graph to fundamentally address cross-domain interference and characterize clearer domain boundaries. By focusing on masked gene expression reconstruction, MMSpa learns stable latent representations that capture core biological features, facilitating the identification of similar spatial subdomains and detecting domain differences across biological samples at the same developmental stage. Comparative analyses using ST datasets from multiple ST technologies and platforms demonstrated that MMSpa outperforms existing methods across various accuracy assessments. Notably, MMSpa excels in challenging scenarios involving highly heterogeneous and complex tissues, and can reveal finer-grained functional tissue domains obscured by other methods. This superior capability positions MMSpa as a powerful tool for uncovering new biological insights and compensating for the lack of spatial annotation in histopathology.

Accurately extracting domain-related information from spatial locations and gene expression is a challenge in spatial transcriptomics analysis. This study develops a masked graph attention autoencoder framework called MMSpa that enhances spatial domain identification in spatial transcriptomics by reducing cross-domain interference and clarifying tissue boundaries.