Analysis, Annotation, and Profiling of the Oat Seed Transcriptome

Novel high-throughput next generation sequencing (NGS) technologies are providing opportunities to explore genomes and transcriptomes in a cost-effective manner. To construct a gene expression atlas of developing oat (Avena sativa) seeds, two software packages specifically designed for RNA-seq (Trinity and Velvet/Oases) were employed for de novo assembly of nearly 134 million quality-filtered 100-bp paired-end reads sequenced from RNA extracted at four stages of seed development. Based on the quality-parameters assessed, Velvet/Oases assemblies were more accurate, produced longer transcripts, and contained more putative unique genes. The final assembly, termed dnOSt (de novo Oat Seed transcriptome), is over 55 million nucleotides in length, contained 53,339 transcript isoforms, and was constructed with approximately 43 million reads, which represents an estimated 74.8× sequencing depth with an average transcript length of 1,043 nucleotides. To assess the accuracy and completeness of dnOSt, we investigated the presence of transcripts associated with the biosynthesis of three compounds with health-promoting properties: avenanthramides, tocols (vitamin E), and ß-glucans. Homologs to all investigated genes were present in dnOSt, demonstrating that it is a robust and useful new tool for oat research. Currently, we are independently assembling transcripts from the four sampled time-points for use in characterizing temporal gene expression patterns during the course of seed development. Results of these analyses will also be presented.