Improving Gene Annotation using RNA-Seq and Transcriptome Shotgun Assembly

Recent advances in genome sequencing technologies which allow generating whole genomes in a fast and cost-effective way have resulted in a steady increase of GenBank submissions of genomes for wide variety of organisms. Accurate annotation of the genomes is still a challenge. More often than not, the transcriptome data for new genomes is available in the form of RNA-Seq which is currently the most cost-effective way of sequencing. The RNA-Seq data available from the deep-sequencing projects has a tremendous potential for the gene finding. Also it creates new challenges for the gene finding algorithms. To name a few they include the sheer amount of the  raw data, a large number of false positive hits and logical problems of finding the right chaining path for connection of small pieces into a full length gene.  In my talk I will describe the method of RNA-Seq data integration adopted in the NCBI eukaryotic genome annotation pipeline which deals with the above challenges and also provides a seamless integration of RNA-Seq and traditional longer transcripts.

Most of the newly sequenced genomes are submitted as draft highly fragmented assemblies. According to our estimates for some organisms thousands of genes are at least partially mapped on missing genome sequence and can’t be fully annotated using the genome sequence. In many cases the missing gene parts could be identified in the Transcriptome Shotgun Assemblies (TSA). I will describe how this information could be incorporated in the gene annotation.