Prediction of Gene Structures Using Public Sequencing Data of RNA-Seq and cDNA in Solanum lycopersicum

To facilitate basic and applied research on Solanum lycopersicum (tomato), comprehensive experimental resources and omics information have been developed and made available. For instance, mutant lines and cDNA clones from a dwarf cultivar, Micro-Tom, are two of the experimental resources. Sequencing data for ESTs and full-length cDNAs (FL-cDNA) from Micro-Tom continue to be accumulated. In conjunction with the reference genome sequence of another cultivar, Heinz 1706, the experimental resources have facilitated comprehensive analyses. To assist in exploiting the experimental resources and omics information, we have maintained a web database, TOMATOMICS (http://bioinf.mind.meiji.ac.jp/tomatomics/). In the TOMATOMICS, information on the experimental resources and omics data of tomato are provided in an integrative manner. Public RNA sequencing (RNA-Seq) data were also integrated as evidence for transcription.

To enhance usability of the transcriptional evidence, we inferred gene structure by using the sequencing data of Heinz 1706 RNA-Seq and Micro-Tom FL-cDNAs. The FL-cDNA sequences were mapped onto the reference genome, then the result was merged with the major gene structural model ITAG2.4. The merged information was used as a supervisor in assembly of RNA-Seq reads into transcripts. The transcripts were applied for prediction of open reading frame, resulted in 28,796 protein-coding genes predicted. The resulting gene structures contained more than 400 of novel gene loci. Furthermore, statistical characterization showed that prediction of untranslated regions and transcript variants were likely improved compared with the ITAG2.4. The information on the now structural annotation, named TMCS version 1.2.1, is also provided in the TOMATOMICS.