scResolve: Recovering single cell expression profiles from multi-cellular spatial transcriptomics

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Current spatial transcriptomics methods often lack single-cell resolution, measuring pooled gene expression from multiple cells.
• This limitation hinders detailed analysis of cellular heterogeneity and cell type-specific functions within complex tissues.

Approach:

• Developed scResolve, a computational method to reconstruct single-cell expression profiles from bulk spatial transcriptomics data.
• scResolve deconvolutes mixed signals to assign gene expression to individual cells at their spatial locations.

Key Points:

• scResolve accurately restores cellular-level gene expression, surpassing traditional cell type deconvolution.
• Successfully applied to human breast cancer and lung disease datasets.
• Enables cell type-specific differential gene expression analysis across different tissue contexts.
• Facilitates accurate identification of rare cell populations.

Conclusions:

• scResolve provides spatially resolved, single-cell expression data, enhancing analytical flexibility.
• Complements existing multi-cellular spatial transcriptomics analyses.
• Opens new avenues for high-resolution spatial biology research.