Improved reconstruction of transcripts and coding sequences from RNA-seq data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Accurate gene and transcript annotation is crucial for understanding newly sequenced genomes.
• RNA-sequencing (RNA-seq) data mapped to genomes provides valuable information for transcript reconstruction.
• Identifying coding sequences within transcripts is essential for understanding protein functions and phenotypes.

Purpose of the Study:

• To present GeMoSeq, a novel computational approach for transcript reconstruction and coding sequence prediction using RNA-seq data.
• To improve the accuracy of gene and transcript annotation, especially for newly sequenced genomes.
• To facilitate the initial annotation of genomes with protein-coding genes.

Main Methods:

• GeMoSeq combines combinatorial enumeration of candidate transcripts with heuristics for exon/intron identification.
• It employs likelihood-based quantification for transcript model evaluation.
• Coding sequence prediction is integrated within the GeMoSeq algorithm.

Main Results:

• GeMoSeq demonstrated improved prediction performance compared to previous methods across seven species.
• Performance gains were particularly notable for coding sequence prediction and in species with dense genomes.
• Re-annotation of Nicotiana benthamiana genomes using GeMoSeq and GeMoMa showcased its utility for initial genome annotation.

Conclusions:

• GeMoSeq offers a robust and accurate method for transcript reconstruction and coding sequence annotation from RNA-seq data.
• The approach is particularly beneficial for annotating newly sequenced genomes, especially those with complex genomic structures.
• GeMoSeq advances the field of genome annotation by providing a comprehensive tool for gene discovery.