Enhanced clustering-based differential expression analysis method for RNA-seq data.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• RNA-sequencing (RNA-seq) is crucial for measuring gene expression and identifying differentially expressed genes (DEGs).
• Gene clustering is typically used for classifying DEGs, not identifying them.
• Previous clustering-based methods (MBCdeg1 and MBCdeg2) showed potential for DEG identification but required enhancement.

Purpose of the Study:

• To introduce an improved clustering-based method, MBCdeg3, for both identifying and classifying DEGs from RNA-seq count data.
• To evaluate the performance of MBCdeg3 against conventional methods like edgeR, DESeq2, and TCC, as well as earlier MBCdeg versions.
• To provide a user-friendly R function for DEG analysis in the expression analysis field.

Main Methods:

• Comparison of six methods: edgeR, DESeq2, TCC, MBCdeg1, MBCdeg2, and MBCdeg3.
• MBCdeg versions utilize different normalization algorithms.
• Performance evaluation using various simulation scenarios of RNA-seq count data.

Main Results:

• MBCdeg3 demonstrates strong performance across diverse simulation scenarios for RNA-seq count data.
• MBCdeg3 effectively identifies and classifies DEGs, outperforming previous versions and some conventional methods in specific contexts.
• The method is implemented as an R function, facilitating its adoption in expression analysis.

Conclusions:

• MBCdeg3 represents a significant improvement for DEG identification and classification from RNA-seq data.
• The R implementation makes MBCdeg3 readily accessible for researchers in gene expression analysis.
• MBCdeg3 offers a reliable and effective tool for analyzing RNA-seq count data in various experimental settings.