Bulk tissue cell type deconvolution with multi-subject single-cell expression reference.

Area of Science:

• Computational biology
• Genomics
• Translational medicine

Background:

• Understanding cell type composition in tissues is crucial for identifying disease targets.
• Bulk RNA sequencing data is abundant but lacks single-cell resolution.
• Single-cell RNA sequencing provides detailed cell-type specific gene expression but is less accessible.

Purpose of the Study:

• To present MuSiC, a computational method for inferring cell type composition from bulk RNA sequencing data.
• To enable the transfer of cell-type specific gene expression information between datasets.
• To leverage large-scale bulk RNA sequencing data for disease mechanism research.

Main Methods:

• MuSiC utilizes cell-type specific gene expression profiles from single-cell RNA sequencing (scRNA-seq).
• It employs weighting strategies for genes with cross-subject and cross-cell consistency.
• The method analyzes bulk RNA sequencing data from complex tissues.

Main Results:

• MuSiC accurately characterizes cell type compositions in bulk RNA-seq data.
• It outperforms existing methods, particularly in tissues with similar cell types (e.g., pancreatic islets, kidney).
• Demonstrated effectiveness across human, mouse, and rat datasets.

Conclusions:

• MuSiC facilitates the characterization of cellular heterogeneity in complex tissues.
• It enables the utilization of vast bulk RNA sequencing resources to study cell type contributions in disease.
• This method advances the identification of cellular targets for disease intervention.