SpaCross: Deciphering Spatial Domains and Integrating Multiple Slices in Spatially Resolved Transcriptomics

Spatially Resolved Transcriptomics (SRT) has revolutionized tissue architecture analysis by integrating gene expression with spatial coordinates. However, existing spatial domain identification methods struggle with unsupervised learning constraints, lack of implicit supervision in latent space, and challenges in balancing local spatial continuity with global semantic consistency, particularly in multi-slice integration. To address these issues, we propose SpaCross, a comprehensive analytical framework for SRT that enhances spatial pattern recognition and cross-slice consistency. SpaCross employs a cross-masked graph autoencoder to reconstruct gene expression features while preserving spatial relationships and mitigating identity mapping issues. A Cross-Masked Latent Consistency module reinforces implicit constraints on latent representations, improving feature robustness. More importantly, an adaptive spatial-semantic graph structure dynamically integrates local and global contextual information, enabling effective multi-slice integration. Extensive evaluations demonstrate that SpaCross outperforms eight state-of-the-art methods on single-slice datasets and achieves robust batch effect correction while preserving biologically meaningful spatial architectures in multi-slice integration. Notably, SpaCross uncovers dynamic spatiotemporal patterns in developing mouse embryos and enables cross-platform integration of mouse olfactory bulb data, detecting shared laminar structures and platform-specific substructures validated by marker gene colocalization. These results highlight its potential to advance spatial transcriptomics analysis in complex tissue systems.