Comparison of Annotation Methods for Comprehensive Crustacean Transcriptome Functional Analysis

Next generation sequencing platforms and improved computational resources can be used to identify novel genes and pathways in a non-model organism. Although there are multiple resources available for de novo assembly, annotation, and measurement of differential expression of RNAseq data, this bioinformatics field is still evolving to generate a standard method of analyzing data. Our goal was to evaluate the current tools and databases available for a comprehensive and comparative annotation of the molting gland, or Y-organ, transcriptome of the blackback land crab, Gecarcinus lateralis. Following assembly using Trinity and sequence clustering via CD-HIT-EST, a transcriptome with 229,278 contigs were generated. We have used multiple resources and databases to assign functional and pathway annotation to sequences. Both nucleotide and protein sequences, obtained via Transdecoder, were used to BLAST against the NCBI non-redundant, Swissprot, and Uniprot-Uniref90 databases that resulted in annotation of 20% of the sequences. The functional annotation of the sequences was performed using Blast2GO and the pathway annotation was generated via Kyoto Encyclopedia of Genes and Genomes. In addition to assigning BLAST information to nucleotide and protein sequences, Trinotate suite was used to identity protein domains, transmembrane domains, and presence of signal peptide sequences. Furthermore, we have used Interproscan resource to match the protein sequences against multiple protein signature databases. A comparison of the above methods is necessary to analyze and predict a standard manner for comprehensive annotation of sequences generated from non-model organisms. Supported by NSF (IOS-1257732).