clrDV: a differential variability test for RNA-Seq data based on the skew-normal distribution.

Area of Science:

• Genomics and Bioinformatics
• Statistical Genetics
• Computational Biology

Background:

• Pathological conditions can alter gene expression variance compared to controls.
• Identifying genes with differential expression variance is crucial for therapeutic target discovery.
• Current methods for differential variability testing in RNA-Seq data are challenged by mean-variance dependence in negative binomial models.

Purpose of the Study:

• To introduce clrDV, a novel statistical method for identifying genes exhibiting differential variability between two populations.
• To address limitations in current RNA-Seq differential variability analysis.

Main Methods:

• Developed clrDV, a statistical method for detecting differential variability in gene expression.
• Utilized the skew-normal distribution for modeling gene-wise null distributions of centered log-ratio transformed compositional RNA-Seq data.

Main Results:

• clrDV demonstrates competitive or superior performance in controlling false discovery rate and Type II error compared to existing methods.
• clrDV offers faster computation times than comparable methods, with performance stable across increasing sample sizes.
• Application of clrDV to a neurodegenerative disease RNA-Seq dataset successfully identified known Alzheimer's disease-associated genes.

Conclusions:

• clrDV is an effective and efficient statistical tool for detecting differential gene expression variability in RNA-Seq data.
• The method shows promise for identifying novel therapeutic targets in complex diseases like Alzheimer's.