STARCH: copy number and clone inference from spatial transcriptomics data.

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Tumors exhibit significant heterogeneity due to diverse cell populations with varying transcriptional and genetic profiles.
• Spatial organization of these cell populations creates a unique tumor microenvironment.
• Copy number aberrations (CNAs) are common genetic alterations in tumors, driving cancer progression by affecting oncogenes and tumor suppressors.

Purpose of the Study:

• To introduce a novel computational method, STARCH (Spatial Transcriptomics Algorithm Reconstructing Copy-number Heterogeneity), for inferring CNAs from spatial transcriptomics data.
• To address the challenges associated with accurately identifying CNAs from RNA sequencing data, particularly in the context of tumor heterogeneity.

Main Methods:

• Spatial transcriptomics technology was utilized to capture gene expression patterns across a grid of spots within tumor tissues.
• The STARCH algorithm was developed to infer CNAs by integrating spatial information, assuming that nearby cells likely share similar CNAs.
• Performance was evaluated against existing methods for CNA inference from RNA sequencing data.

Main Results:

• STARCH effectively infers copy number aberrations from spatial transcriptomics data.
• The method leverages the spatial proximity of cells to improve the accuracy of CNA detection.
• STARCH demonstrates superior performance compared to non-spatial RNA sequencing-based methods for CNA inference.

Conclusions:

• STARCH provides a powerful new tool for dissecting tumor heterogeneity at the genomic level using spatial transcriptomics.
• Accurate inference of CNAs from spatial data can enhance our understanding of tumor evolution and microenvironment.
• This approach holds promise for advancing cancer research and potentially guiding therapeutic strategies.