Despite a substantial investment in the development of panels of single nucleotide polymorphism (SNP) markers, the simple sequence repeat (SSR) technology with a limited multiplexing capability remains a standard, even for applications requiring whole-genome information. Diversity arrays technology (DArT) types hundreds to thousands of genomic loci in parallel, as previously demonstrated in a number diploid plant species. Here we show that DArT performs similarly well for the hexaploid genome of bread wheat (Triticum aestivum L.). The methodology previously used to generate DArT fingerprints of barley also generated a large number of high-quality markers in wheat (99.8% allele-calling concordance and approximately 95% call rate). The genetic relationships among bread wheat cultivars revealed by DArT coincided with knowledge generated with other methods, and even closely related cultivars could be distinguished. To verify the Mendelian behaviour of DArT markers, we typed a set of 90 Cranbrook x Halberd doubled haploid lines for which a framework (FW) map comprising a total of 339 SSR, restriction fragment length polymorphism (RFLP) and amplified fragment length polymorphism (AFLP) markers was available. We added an equal number of DArT markers to this data set and also incorporated 71 sequence tagged microsatellite (STM) markers. A comparison of logarithm of the odds (LOD) scores, call rates and the degree of genome coverage indicated that the quality and information content of the DArT data set was comparable to that of the combined SSR/RFLP/AFLP data set of the FW map.

Understanding the distribution of genetic diversity among individuals, populations and gene pools is crucial for the efficient management of germplasm collections and breeding programs. Diversity analysis is routinely carried out using sequencing of selected gene(s) or molecular marker technologies. Here we report on the development of Diversity Arrays Technology (DArT) for pigeonpea (Cajanus cajan) and its wild relatives. DArT tests thousands of genomic loci for polymorphism and provides the binary scores for hundreds of markers in a single hybridization-based assay. We tested eight complexity reduction methods using various combinations of restriction enzymes and selected PstI/HaeIII genomic representation with the largest frequency of polymorphic clones (19.8%) to produce genotyping arrays. The performance of the PstI/HaeIII array was evaluated by typing 96 accessions representing nearly 20 species of Cajanus. A total of nearly 700 markers were identified with the average call rate of 96.0% and the scoring reproducibility of 99.7%. DArT markers revealed genetic relationships among the accessions consistent with the available information and systematic classification. Most of the diversity was among the wild relatives of pigeonpea or between the wild species and the cultivated C. cajan. Only 64 markers were polymorphic among the cultivated accessions. Such narrow genetic base is likely to represent a serious impediment to breeding progress in pigeonpea. Our study shows that DArT can be effectively applied in molecular systematics and biodiversity studies.

Understanding the distribution of genetic diversity within and among individuals, populations, species and gene pools is crucial for the efficient management of germplasm collections. Molecular markers are playing an increasing role in germplasm characterization, yet their broad application is limited by the availability of markers, the costs and the low throughput of existing technologies. This is particularly true for crops of resource-poor farmers such as cassava, Manihot esculenta. Here we report on the development of Diversity Arrays Technology (DArT) for cassava. DArT uses microarrays to detect DNA polymorphism at several hundred genomic loci in a single assay without relying on DNA sequence information. We tested three complexity reduction methods and selected the two that generated genomic representations with the largest frequency of polymorphic clones (PstI/TaqI: 14.6%, PstI/BstNI: 17.2%) to produce large genotyping arrays. Nearly 1,000 candidate polymorphic clones were detected on the two arrays. The performance of the PstI/TaqI array was validated by typing a group of 38 accessions, 24 of them in duplicate. The average call rate was 98.1%, and the scoring reproducibility was 99.8%. DArT markers displayed fairly high polymorphism information content (PIC) values and revealed genetic relationships among the samples consistent with the information available on these samples. Our study suggests that DArT offers advantages over current technologies in terms of cost and speed of marker discovery and analysis. It can therefore be used to genotype large germplasm collections.