Computational analysis in spatial transcriptomics: methods and perspectives

Spatial transcriptomics (STs) enables spatially resolved gene-expression profiling across diverse tissues, generating large-scale datasets that integrate molecular and spatial information. To analyze and interpret the ST data, a wide range of computational methods have been proposed. As a result of continued expansion in the quantity and complexity of the methods for ST analysis, there is an increasing need for clear and structured guidance to help researchers effectively apply appropriate strategies in their study. In this review, we present a comprehensive overview of the current state of computational approaches in ST data analysis, covering key aspects concerning data storage, data preprocessing, resolution enhancement, and downstream analyses, including spatial domain identification, spatially variable genes detection, cell type annotation, cell-cell communication, gene expression prediction modeling, and 3D ST reconstruction. Across these areas, we summarize recent methodological advances, highlight remaining challenges, and outline future directions for advancing ST analysis, particularly in computational methods related to high-resolution ST platforms.