CellMapper: rapid and accurate inference of gene expression in difficult-to-isolate cell types

Area of Science:

• Genomics
• Bioinformatics
• Cell Biology

Background:

• Identifying cell type-specific gene expression is crucial for understanding cellular function and disease.
• Existing computational methods face challenges, particularly for rare or unisolated cell types.
• Publicly available gene expression datasets offer a rich resource for computational analysis.

Purpose of the Study:

• To develop and validate a sensitive computational approach for predicting cell type-specific gene expression.
• To outperform existing algorithms in accuracy, especially for challenging cell types.
• To demonstrate a clinically relevant application in prioritizing genes from genome-wide association studies (GWAS).

Main Methods:

• CellMapper utilizes publicly available gene expression data.
• The algorithm searches for genes with expression profiles similar to known cell-specific markers.
• The method was evaluated against previous computational algorithms and validated on diverse cell types.

Main Results:

• CellMapper significantly outperforms existing computational methods for predicting cell type-specific gene expression.
• The approach accurately identifies expression patterns for rare and difficult-to-isolate cell types.
• Accurate predictions were achieved for human brain cell types, including those not yet isolated.
• The method demonstrated successful application in prioritizing candidate genes within GWAS-identified disease susceptibility loci.

Conclusions:

• CellMapper provides a sensitive and accurate method for predicting cell type-specific gene expression.
• The tool is broadly applicable across various tissues and cell types, including rare ones.
• CellMapper has significant potential for advancing basic research and clinical applications, such as disease gene discovery.