Reannotation Of The Bovine Genome Using RNA Sequencing Data

The availability of the bovine genome sequence assembly and haplotype map marked a significant development in the application of genomics for the genetic improvement and biological understanding of cattle.  As a next step in cattle genomics, it is essential to enhance functional annotation of the reference sequence relevant to the biology of the species and for comparative analysis with humans and other species. Current bovine gene annotations are primarily based on computationally predicted models and therefore contain inaccuracies and omissions. To enhance the value of the reference assembly to cattle genomics and dairy and beef sciences a more accurate and complete gene annotation is essential.  Deep sequencing technology makes it possible to generate large amounts of transcriptome sequence data that can be used to improve genome annotation. In this study, two billion uniquely mapped reads from RNASeq experiments representing 12 tissue types, 26 individuals and 5 breeds were used to extend the Ensembl UMD3.1.68 annotation. Analysis of these data contributed 5,846 new genes, new alternative transcripts for 10,941 genes, and extended 3’ and 5’ sequences for 4,640 and 4,379 genes, respectively. The resulting improved annotation will contribute to a more comprehensive analysis and better interpretation of cattle genomic and genetic studies.