STModule: identifying tissue modules to uncover spatial components and characteristics of transcriptomic landscapes

Area of Science:

• Spatial transcriptomics
• Computational biology
• Bioinformatics

Background:

• Spatially resolved transcriptomics provides insights into tissue architecture and cellular interactions.
• Identifying distinct tissue modules is crucial for understanding complex biological systems.
• Existing methods may not capture the full spectrum of biological signals in transcriptomic landscapes.

Purpose of the Study:

• To develop and present STModule, a novel Bayesian method for identifying tissue modules from spatial transcriptomics data.
• To reveal spatial components and essential characteristics of various tissues, including those in cancer.
• To facilitate downstream analysis and provide deeper insights into tumor microenvironments and disease mechanisms.

Main Methods:

• STModule employs a Bayesian approach to analyze spatially resolved transcriptomic data.
• The method identifies tissue modules by uncovering diverse expression signals.
• Gene sets characterize the identified tissue modules.

Main Results:

• STModule successfully uncovers diverse expression signals in transcriptomic landscapes like cancer and immune infiltrates.
• The method detects novel spatial components and captures a broader spectrum of biological signals than other approaches.
• Characterized tissue modules demonstrate enhanced robustness and transferability across different biopsies.

Conclusions:

• STModule is an effective tool for identifying tissue modules from spatial transcriptomics data.
• The identified modules offer valuable insights into tumor microenvironments, disease mechanisms, and histological organization.
• STModule enhances downstream analysis and provides a more comprehensive understanding of tissue biology.