Scabiosa columbaria Transcriptome or Gene expression

Background: Next-generation sequencing (NGS) technologies are increasingly applied in many organisms, including non-model organisms that are important for ecological and conservation purposes. Usually, the choice among different NGS technologies depends on a trade-off between the available funds and the desired depth of information. Illumina and 454 sequencing are among the most used NGS technologies and have been predicted to produce optimal results at reasonable costs when used together. This combination has been successfully applied to study the genome of bacteria and fungi, but not yet to characterize the transcriptome of non-model plant species. Here we describe the combined application of these two NGS technologies to a plant species with ecological and conservation relevance for which to date no genomic resources is available, Scabiosa columbaria. In particular, we evaluated how 454 GS-FLX and Illumina GAII reads perform, both in combination and individually, in producing high-quality de novo assemblies of gene transcripts. We then used the outcome of our highest quality de novo assembly to further characterize the transcriptome of S. columbaria. Results: We obtained 528,557 reads (mean length 212 bp) from a 454 GS-FLX run and a total of 28,993,627 reads (all of 75 bp) from two lanes of an Illumina GAII single run. After reads trimming, the de novo assembly of Illumina reads produced 103,500 contigs, while the de novo assembly of 454 reads produced 36,818 contigs. De novo assembly including both types of trimmed reads produced 109,630 contigs and was shown to be the highest-quality assembly among the three applied. Based on sequence similarity with known gene products, these sequences represent at least 12,516 unique genes, most of which are well covered by contig sequences. Conclusions: This study empirically demonstrates that the combination of 454 pyrosequencing and Illumina sequencing technologies can produce a higher-quality de novo assembly when compared to their individual use. In addition, the resulting sequence collection represents a main genomic resource for S. columbaria which should contribute to future research in conservation and population biology studies. Our results demonstrate the utility of NGS technologies as starting point for the development of genomic tools in non-model but ecologically important species.