Clustering of Differentially Expressed Genes from Transcriptome of Vitis flexuosa

Transcriptome analysis is one of powerful tools to select valuable genes and genetic information in grape breeding program. In the present study, transcriptome of flower (full blooming and 7 days before flowering), leaf, fruit (young green, and ripe fruit), and root from Vitis flexuosa was analyzed to select useful genes, to elucidate their function, and to compare their differential expression through assembly, selection of DEGs, clustering, and annotation (GO and KEGG) of data from sequencing short reads on Solexa platform. We have assessed the effect of sequence quality, various assembly parameters and assembly programs on the final assembly output. We assembled ~132 million high-quality trimmed reads using Velvet followed by Oases with optimal parameters into a non-redundant set of 188,058 transcripts (≥ 100 bp in length), representing about 41 Mb of unique transcriptome sequence. The average length of transcripts was 1,722 bp and N50 length of 2,182 bp with largest contig length of 12,228 bp. Among assembles transcripts, a total of 31,834 V. flexuosa transcripts were selected as unigenes/predicted proteins from sequenced V. vinifera or other plant  genomes at the protein level. From them, 143 unique loci were selected sfecifically from V. flexuosa based on similarity with V. vinifera and other plant genomes. Functional categorization revealed the conservation of genes involved in various biological processes like primary metabolic process (33.3%), cellular metabolic process (32.3%), and cellular metabolic process (33.3%) in V. flexuosa. The V. flexuosa transcripts set generated here will provide a resource for gene discovery and development of functional molecular markers. In addition, the strategy for assembly of transcriptome data presented here will be helpful in other similar transcriptome studies.