De novo assembly of unpaired sequencing reads yields contig breaks at repeat sequences that are longer than the sequencing read, e.g. transposable elements, rRNA operons, and tRNA clusters.
Our results demonstrate that de novo assembly of short (Illumina) reads yields high-quality, high-coverage transcriptomic data sets.
Here, we demonstrate that the sole use of PacBio reads yields a larger genome assembly with fewer gaps than similar genomes generated through capillary sequencing or 454 sequencing.
Additionally, low DNA input libraries for 454 (linker-amplified [ 15] to obtain sufficient genetic material) were all successful, with highest read yields per ng of input DNA of any method (Additional file 1: Figure S1).
Assembly of 90% of the shotgun reads yielded 4938 contigs with an average length of 2752 bp, which were further assembled with 20% of the fosmid library sequence reads, generating 655 contigs with average length of 6273 bp.
Our results suggest that fluctuations in read yield and sequence quality are more likely driven by individual run performance than by input template amount.
A total of 37,533,509 smRNA reads yielded 11,129 independent orchid miRNA sequences, representing 329 known miRNA families identified in other plant species.
Filtering low quality reads yielded ∼35.9 million reads, and ∼23.0 million (64%) of these were aligned to the genome with at most two mismatches in order to cope with sequencing errors and SNPs.
We did not observe a clear relationship between input template DNA concentration and 454 sequence read yield or quality; evaluation of the total number of reads relative to the number expected versus input template quantity did not show a template dependent difference in the read ratio.
The filtered reads yield approximately 100 Mb of quality data.
