The advent of the second generation sequencing technology has made RNA sequencing (RNA-Seq) a preferable choice over existing transcriptome profiling methods. As RNA-seq experiments can be designed to output data which do not require a reference genome to analyze, it is a cost efficient and high-throughput choice for transcriptome profiling of non-model organisms which only have partial, if not any reference genome. However, as many second generation sequencers including the popular Illumina system are not capable of reading the full length of many transcripts, the reads must first be assembled in effort to reconstruct the original transcriptome. There have been studies independently addressing the influences of some sequencing parameters in de novo transcriptome assembly; namely the read depth and read length. Both reports suggest the existence of a fundamental limit for performance enhancement, but to date there is no study which comprehensively analyzes the influences of such parameters in actual RNA-seq data which are publicly accessible. In this research, RNA-seq data for four model organisms were obtained from the sequencing read archive (SRA), categorized by read depth and read length, assembled using the SOAPdenovo-Trans and Trinity software, and evaluated with several assembly metrics and by searching against a well defined subset of Clusters of Orthologous Groups (COGs) included in the Core Eukaryotic Genes Mapping Approach (CEGMA) pipeline. By studying these results we aimed to generate guidelines to selecting sequencing parameters for RNA-seq experiments targeted to non-model organisms. However, assembly results were not consistent most likely due to lack of data.

慶應義塾大学湘南藤沢キャンパス先端生命科学研究会 2014年度学生論文集
卒業論文ダイジェスト