Grouper: graph-based clustering and annotation for improved de novo transcriptome analysis.

Area of Science:

• Transcriptomics
• Bioinformatics
• Computational Biology

Background:

• De novo transcriptome analysis via RNA-sequencing is crucial for gene expression studies in non-model organisms.
• Transcriptome assembly often yields fragmented contigs, hindering robust downstream analyses.
• Existing methods struggle to accurately group related transcripts and genes from fragmented assemblies.

Purpose of the Study:

• To introduce Grouper, a novel computational method for clustering de novo assembled transcriptome contigs.
• To enhance the accuracy and completeness of transcriptome assemblies for improved gene expression studies.
• To facilitate robust downstream analyses, including differential gene expression.

Main Methods:

• Grouper employs a clustering algorithm to group contigs likely originating from the same transcripts or genes.
• The method can optionally utilize a related annotated genome to transfer annotations and refine clustering.
• Evaluated on de novo assemblies from four diverse species.

Main Results:

• Grouper accurately clusters a significantly larger number of contigs compared to state-of-the-art methods.
• The Grouper pipeline improves read mapping to contigs by over 10%, enhancing differential expression analysis accuracy.
• Annotation transfer from related genomes effectively improves clustering results.

Conclusions:

• Grouper provides a complete and efficient pipeline for processing de novo transcriptomic assemblies.
• The software enhances the reliability of gene expression studies in non-model organisms.
• Grouper is freely available, promoting wider adoption in the research community.