Corset: enabling differential gene expression analysis for de novo assembled transcriptomes.

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Next-generation sequencing enables differential gene expression studies, even without a reference genome.
• De novo transcriptome assembly from RNA-seq data generates numerous contigs requiring gene-level clustering for analysis.

Purpose of the Study:

• To introduce Corset, a novel method for hierarchical clustering of RNA-seq contigs.
• To prepare clustered contigs for differential gene expression detection by summarizing read counts.

Main Methods:

• Corset employs hierarchical clustering based on shared reads and expression patterns.
• The method summarizes read counts to the gene cluster level.

Main Results:

• Corset demonstrates superior performance compared to existing alternative methods across various metrics.
• The clustering approach facilitates downstream statistical analysis for gene expression.

Conclusions:

• Corset provides an effective solution for gene clustering in de novo transcriptome assemblies.
• This method enhances the analysis of differential gene expression in non-model organisms.