GASS: genome structural annotation for Eukaryotes based on species similarity

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• High-throughput sequencing has increased genome assembly, but annotation remains challenging.
• Eukaryotic genome annotation requires extensive, diverse data, making the process complex and time-consuming.

Purpose of the Study:

• To develop a fast and user-friendly computational tool for eukaryotic genome annotation.
• To enable rapid genome annotation using data from only a similar species.

Main Methods:

• Proposed GASS (Genome Annotation based on Species Similarity), a tool utilizing annotations from a similar species.
• Employed a shortest-path model to align exons and detect optimal transcript annotations.
• Applied GASS to rhesus genome annotation using human annotations.

Main Results:

• GASS achieved over 65% exact matches for RefSeq exons and splicing junctions in rhesus.
• Demonstrated higher sensitivity than RefSeq and comparable sensitivity to Ensembl.
• Exhibited superior specificity across gene, transcript, exon, and splicing junction levels compared to Ensembl.
• Identified potential mis-assemblies in the rheMac3 genome impacting annotations.

Conclusions:

• GASS rapidly generates accurate and abundant structural genome annotations.
• Empowers smaller laboratories to obtain timely genome annotations without extensive data requirements.
• Facilitates analysis of RNA-Seq data for species with available genome drafts but lacking annotations.