Computational analysis in spatial transcriptomics: methods and perspectives

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Spatial transcriptomics (STs) generates large-scale datasets integrating molecular and spatial information from diverse tissues.
• The increasing volume and complexity of ST data necessitate structured guidance for computational analysis.

Purpose of the Study:

• To provide a comprehensive overview of current computational approaches for ST data analysis.
• To guide researchers in selecting appropriate strategies for their ST studies.
• To summarize methodological advances, challenges, and future directions in ST analysis.

Main Methods:

• Review of computational methods for ST data analysis.
• Coverage of data storage, preprocessing, and resolution enhancement.
• Inclusion of downstream analyses: spatial domain identification, spatially variable genes detection, cell type annotation, cell-cell communication, gene expression prediction modeling, and 3D ST reconstruction.

Main Results:

• Summarizes recent advances in computational methods across key ST analysis areas.
• Identifies remaining challenges in ST data interpretation.
• Outlines future research directions, with a focus on high-resolution ST platforms.

Conclusions:

• A structured overview of computational ST analysis is crucial for advancing the field.
• Continued development of computational methods is needed, especially for high-resolution and 3D ST data.
• This review serves as a guide for researchers navigating the evolving landscape of ST data analysis.