Landscape of the Dark Transcriptome Revealed Through Re-mining Massive RNA-Seq Data

Area of Science:

• Genomics and Transcriptomics
• Molecular Biology
• Bioinformatics

Background:

• The "dark transcriptome" comprises transcribed sequences lacking gene annotation.
• Understanding these unannotated regions is crucial for a complete view of the genome's functional potential.

Purpose of the Study:

• To investigate the expression and potential protein-coding capacity of unannotated open reading frames (ORFs) in the Saccharomyces cerevisiae genome.
• To identify candidate protein-coding genes within the dark transcriptome.
• To provide a framework for exploring and reusing public RNA-Seq data for novel gene discovery.

Main Methods:

• Analyzed 3,457 RNA-Seq samples from Saccharomyces cerevisiae under diverse conditions.
• Evaluated expression of 6,692 annotated genes and 29,354 unannotated ORFs.
• Utilized phylostratigraphic analysis and Markov Chain Clustering for ORF classification and co-expression analysis.
• Integrated data into MetaOmGraph (MOG) for interactive visualization and analysis.

Main Results:

• Over 30% of highly transcribed ORFs demonstrated translation evidence.
• Phylostratigraphic analysis suggested most transcribed ORFs encode species-specific proteins ("orphan-ORFs").
• Hundreds of orphan-ORFs exhibited expression levels comparable to annotated genes.
• Markov Chain Clustering identified 2,468 orphan-ORFs within co-expression clusters.

Conclusions:

• A significant portion of the dark transcriptome in yeast likely represents unannotated protein-coding genes.
• These findings highlight the potential for novel gene discovery within previously overlooked genomic regions.
• The MetaOmGraph platform facilitates the exploration of large-scale RNA-Seq data, enabling hypothesis generation for experimental validation.