Meeting Abstract

41.4  Wednesday, Jan. 5  Cross validation of next-generation sequencing platforms for gene expression studies in non-model organisms SIEBERT, S.*; ROBINSON, M.; SMITH, S.A.; TINTORI, S.C.; HELM, R.R.; GOETZ, F.; HADDOCK, S.H.; CARTWRIGHT, P.; DUNN, C.W.; Brown University, Providence; Walter and Eliza Hall Institute of Medical Research, Melbourne; Brown University, Providence; Brown University, Providence; Brown University, Providence; Brown University, Providence; Monterey Bay Aquarium Research Institute, Moss Landing; University of Kansas, Lawrence; Brown University, Providence stefan_siebert@brown.edu

Next generation sequencing has become a critical tool in studies of gene expression in model organisms. In this application, the number of transcriptome sequence fragments that map to each gene indicates expression level. The primary hurdle to using these tools with non-model species is deriving the gene predictions without a complete genome, and in understanding what potential biases may exist when mapping to such a partial reference. Here we use a hybrid sequencing strategy to investigate differential gene expression between functionally specialized polyps and medusae in the deep-sea siphonophore Nanomia bijuga (Cnidaria). We combine long-read sequencing (Roche 454) for gene sequence prediction with short-read sequencing for expression quantification. In order to cross-validate expression quantification, we used three different short read sequencing platforms — Illumina, SOLiD, and Helicos. We find congruences across platforms, complementary additional value and explore particular cases in which they depart. These findings verify that robust functional genomics analyses are now possible in non-model species, and provide specific insight into the optimization of sequencing-based gene expression studies in model and non-model species alike.