BackgroundLinkage maps are an integral resource for dissection of complex genetic traits in plant and animal species. Canonical map construction follows a well-established workflow: an initial discovery phase where genetic markers are mined from a small pool of individuals, followed by genotyping of selected mapping populations using sets of marker panels. A newly developed sequence-based marker technology, Restriction site Associated DNA (RAD), enables synchronous single nucleotide polymorphism (SNP) marker discovery and genotyping using massively parallel sequencing. The objective of this research was to assess the utility of RAD markers for linkage map construction, employing barley as a model system. Using the published high density EST-based SNP map in the Oregon Wolfe Barley (OWB) mapping population as a reference, we created a RAD map using a limited set of prior markers to establish linakge group identity, integrated the RAD and prior data, and used both maps for detection of quantitative trait loci (QTL).ResultsUsing the RAD protocol in tandem with the Illumina sequence by synthesis platform, a total of 530 SNP markers were identified from initial scans of the OWB parental inbred lines - the "dominant" and "recessive" marker stocks - and scored in a 93 member doubled haploid (DH) mapping population. RAD sequence data from the structured population was converted into allele genotypes from which a genetic map was constructed. The assembled RAD-only map consists of 445 markers with an average interval length of 5 cM, while an integrated map includes 463 RAD loci and 2383 prior markers. Sequenced RAD markers are distributed across all seven chromosomes, with polymorphic loci emanating from both coding and noncoding regions in the Hordeum genome. Total map lengths are comparable and the order of common markers is identical in both maps. The same large-effect QTL for reproductive fitness traits were detected with both maps and the majority of these QTL were coincident with a dwarfing gene (ZEO) and the VRS1 gene, which determines the two-row and six-row germplasm groups of barley.ConclusionsWe demonstrate how sequenced RAD markers can be leveraged to produce high quality linkage maps for detection of single gene loci and QTLs. By combining SNP discovery and genotyping into parallel sequencing events, RAD markers should be a useful molecular breeding tool for a range of crop species. Expected improvements in cost and throughput of second and third-generation sequencing technologies will enable more powerful applications of the sequenced RAD marker system, including improvements in de novo genome assembly, development of ultra-high density genetic maps and association mapping.

Barley (Hordeum vulgare) is an important component of heart-healthy whole grain diets because it contains b-glucan. All current US barley varieties with high b-glucan are spring habit and have waxy starch. Winter varieties have agronomic advantages but require low-temperature tolerance (LTT). Vernalization sensitivity (VS) is associated with higher levels of LTT. To rapidly develop fall-sown varieties with LTT and higher grain b-glucan, we therefore used marker-assisted selection (MAS) at the WX and VRN-H2 loci. The MAS-derived lines, together with unrelated non-waxy germplasm developed via phenotypic selection (PS), were used for a genome-wide association scan (GWAS). The panel was phenotyped for grain b-glucan, LTT and VS. It was genotyped with 3072 single-nucleotide polymorphisms (SNPs) and allele-specific primers. Marker-assisted selection fixed target alleles at both loci but only one of the target phenotypes (higher b-glucan percentage) was achieved. Variation for VS and LTT is attributable to (i) incomplete information about VRN-H1 at the outset of the project and (ii) unexpected allelic variation at VRN-H3 with a large effect on VS and LTT.

The Oregon Wolfe Barley mapping population is a resource for genetics research and instruction. Prior reports are based on a population of doubled haploid (DH) lines developed by the Hordeum bulbosum (H.b.) method, which samples female gametes. We developed new DH lines from the same cross using anther culture (A.C.), which samples male gametes. Linkage maps were generated in each of the two subpopulations using the same 1,328 single nucleotide polymorphism markers. The linkage maps based on DH lines derived from the products of megasporogeneis and microsporogenesis revealed minor differences in terms of estimated recombination rates. There were no differences in locus ordering. There was greater segregation distortion in the A.C.-derived subpopulation than in the H.b.-derived subpopulation, but in the region showing the greatest distortion, the cause was more likely allelic variation at the ZEO1 plant height locus rather than to DH production method. The effects of segregation distortion and pleiotropy had greater impacts on estimates of quantitative trait locus effect than population size for reproductive fitness traits assayed under greenhouse conditions. The Oregon Wolfe Barley (OWB) population and data are community resources. Seed is available from three distribution centers located in North America, Europe, and Asia. Details on ordering seed sets, as well as complete genotype and phenotype data files, are available at http://wheat.pw.usda.gov/ggpages/maps/OWB/ .