Regional analysis to delineate intrasample heterogeneity with RegionalST

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Spatial transcriptomics advances understanding of tissue heterogeneity crucial for disease mechanisms.
• Intra-tumor heterogeneity impacts cancer treatment response.
• Current computational tools lack capabilities for cross-regional analysis and have limited spatial resolution.

Purpose of the Study:

• Introduce RegionalST, an efficient computational method for spatial transcriptomics data analysis.
• Enable quantification of cell type mixtures and interactions.
• Facilitate identification of sub-regions and cross-region differential analysis.

Main Methods:

• Developed an efficient computational method named RegionalST.
• Utilized simulations and real-world spatial transcriptomics data for validation.
• Implemented the method as an open-source R/Bioconductor package.

Main Results:

• RegionalST enables quantification of cell type mixtures and interactions.
• The tool facilitates sub-region identification and cross-region differential analysis.
• Demonstrated efficiency and accuracy in visualizing and analyzing diverse spatial transcriptomics data.

Conclusions:

• RegionalST is an efficient, accurate, and flexible tool for exploring tissue heterogeneity.
• Provides a comprehensive solution for in-depth spatial transcriptomics data analysis.
• The open-source package is readily available for researchers.