Using Spinach to Compare Technologies for Whole Genome Assemblies

Spinach (Spinach oleraceae) is a diploid, annual,  dioecious crop (2n=12) bred for three different commodities: fresh market clipped and bagged, fresh market bunched and frozen with a value of $US 7.85 B in 2009 worldwide (Food and Agriculture Organization of the United Nations 2012). Selection for traits depends on market type including leaf shape, bolting resistance, shelf life and disease resistance, particularly against Pernospora farinose-downy mildew, and yield. The vast majority of varieties are hybrids.  Relative to many other crops, spinach lacks genomic resources with relatively few genetic markers for traits. Genetic resources such as maps exist with a 100 AFLP and SSR markers (Khattak et al. 2006).  Publicly, only 225 ESTs exist in GenBank and another 1006 nucleotide sequences, the vast majority making up the sequenced chloroplast genome of 150 kbp. With a genome size of 989 Mb, the sequencing and annotation of spinach is feasible.  We have assembled the genome using a sequence form multiple Illumina libraries. We have also assembled a de novo reference sequence using Pacific Biosciences only and large fragment libraries.  Improved algorithms to assemble long reads were applied showing a 63 fold improvement in contig size over Illumina only assembly. The reference genome will allow rapid development of genetic maps, annotation, expression analyses, and analyses of diversity through resequencing.