Sunetra Das1, Donald L Mykles2. 1. Department of Biology, Colorado State University, Fort Collins, CO 80523, USA. 2. Department of Biology, Colorado State University, Fort Collins, CO 80523, USA donald.mykles@colostate.edu.
Abstract
Next-generation sequencing technologies are revolutionizing crustacean biology. De novo assembly of RNA sequencing (RNA-seq) data allows researchers to catalog and quantify genes expressed in tissues of a species that lacks a complete genome sequence. RNA-seq has become an important tool for understanding phenotypic plasticity and the responses of organisms to environmental cues. However, there are challenges with identification of assembled contiguous sequences (contigs) without a reference genome. Thus, the selection of resources for annotating contigs is critical for the downstream analysis of gene functions. A de novo-assembled transcriptome of the Gecarcinus lateralis molting gland, or Y-organ (YO), was used to compare two functional annotation packages, Trinotate and Blast2GO. The assembled transcriptome contained 229,278 contigs derived from YOs from animals in intermolt, premolt (early, mid, and late), and postmolt stages. Gene identification using BLAST against four databases and functional annotation using Gene Ontologies were conducted. The analysis revealed two major limitations of de novo assembly: (1) assembly using Trinity generates redundant contigs and (2) transcripts that encode protein isoforms are not distinguished with current computational tools. It is recommended that the NCBI Non-Redundant, SwissProt, TrEMBL, and Uniref90 databases be used to maximize gene identification. Trinotate is preferred for assigning functions to identified genes, as the package uses multiple databases for annotation. The differences between packages to generate Gene Ontology (GO) terms are attributed to the databases used for inputs: Trinotate uses both Pfam and BLAST databases, while Blast2GO uses only the BLAST database. InterProScan was used to verify the GO terms assigned via BLAST. A comprehensive annotation of de novo assembled transcriptome is necessary for the downstream analysis of differentially expressed transcripts in the YO over the molt cycle.
Next-generation sequencing technologies are revolutionizing crustacean biology. De novo assembly of RNA sequencing (RNA-seq) data allows researchers to catalog and quantify genes expressed in tissues of a species that lacks a complete genome sequence. RNA-seq has become an important tool for understanding phenotypic plasticity and the responses of organisms to environmental cues. However, there are challenges with identification of assembled contiguous sequences (contigs) without a reference genome. Thus, the selection of resources for annotating contigs is critical for the downstream analysis of gene functions. A de novo-assembled transcriptome of the Gecarcinus lateralis molting gland, or Y-organ (YO), was used to compare two functional annotation packages, Trinotate and Blast2GO. The assembled transcriptome contained 229,278 contigs derived from YOs from animals in intermolt, premolt (early, mid, and late), and postmolt stages. Gene identification using BLAST against four databases and functional annotation using Gene Ontologies were conducted. The analysis revealed two major limitations of de novo assembly: (1) assembly using Trinity generates redundant contigs and (2) transcripts that encode protein isoforms are not distinguished with current computational tools. It is recommended that the NCBI Non-Redundant, SwissProt, TrEMBL, and Uniref90 databases be used to maximize gene identification. Trinotate is preferred for assigning functions to identified genes, as the package uses multiple databases for annotation. The differences between packages to generate Gene Ontology (GO) terms are attributed to the databases used for inputs: Trinotate uses both Pfam and BLAST databases, while Blast2GO uses only the BLAST database. InterProScan was used to verify the GO terms assigned via BLAST. A comprehensive annotation of de novo assembled transcriptome is necessary for the downstream analysis of differentially expressed transcripts in the YO over the molt cycle.
Authors: Donald L Mykles; Karen G Burnett; David S Durica; Blake L Joyce; Fiona M McCarthy; Carl J Schmidt; Jonathon H Stillman Journal: Integr Comp Biol Date: 2016-09-17 Impact factor: 3.326
Authors: Alexey А Moskalev; Anna V Kudryavtseva; Alexander S Graphodatsky; Violetta R Beklemisheva; Natalya A Serdyukova; Konstantin V Krutovsky; Vadim V Sharov; Ivan V Kulakovskiy; Andrey S Lando; Artem S Kasianov; Dmitry A Kuzmin; Yuliya A Putintseva; Sergey I Feranchuk; Mikhail V Shaposhnikov; Vadim E Fraifeld; Dmitri Toren; Anastasia V Snezhkina; Vasily V Sitnik Journal: BMC Evol Biol Date: 2017-12-28 Impact factor: 3.260