Detection of condition-specific marker genes from RNA-seq data with MGFR

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Identifying condition-specific genes is crucial for understanding cell fate and disease.
• Differential gene expression analysis (DGEA) is standard but challenging with large datasets or few replicates.
• Previous tool MGFM was effective for microarrays, especially with limited samples.

Purpose of the Study:

• To adapt the MGFM algorithm for marker gene detection in RNA-seq data, named MGFR.
• To evaluate MGFR's performance against existing methods for marker detection.
• To identify novel candidate marker genes for various human tissues and cell types.

Main Methods:

• MGFR groups samples by gene expression and flags markers based on highest expression across replicates.
• Benchmarking was performed using standard and single-cell RNA-seq datasets.
• Detailed analysis was conducted on human tissue, immune, and blastocyst cell type datasets.

Main Results:

• MGFR markers accurately characterized the functional identity of different tissues and cell types.
• MGFR outperformed other marker detection methods when compared to gold-standard lists.
• Novel candidate marker genes were identified for the studied tissues and cell types.

Conclusions:

• MGFR is an effective tool for marker gene detection in RNA-seq data, particularly for datasets with low replicate numbers.
• The tool accurately identifies functional markers and can suggest novel candidates.
• MGFR is available as a Bioconductor package for easy integration into bioinformatics workflows.