Development of a next generation SNP genotyping array for wheat

Abstract

High-throughput genotyping arrays have provided a cost-effective, reliable and interoperable system for genotyping hexaploid wheat and its relatives. Existing, highly cited arrays including our 35K Wheat Breeder's array and the Illumina 90K array were designed based on a limited amount of varietal sequence diversity and with imperfect knowledge of SNP positions. Recent progress in wheat sequencing has given us access to a vast pool of SNP diversity, whilst technological improvements have allowed us to fit significantly more probes onto a 384-well format Axiom array than previously possible. Here we describe a novel Axiom genotyping array, the 'Triticum aestivum Next Generation' array (TaNG), largely derived from whole genome skim sequencing of 204 elite wheat lines and 111 wheat landraces taken from the Watkins 'Core Collection'. We used a novel haplotype optimization approach to select SNPs with the highest combined varietal discrimination and a design iteration step to test and replace SNPs which failed to convert to reliable markers. The final design with 43 372 SNPs contains a combination of haplotype-optimized novel SNPs and legacy cross-platform markers. We show that this design has an improved distribution of SNPs compared to previous arrays and can be used to generate genetic maps with a significantly higher number of distinct bins than our previous array. We also demonstrate the improved performance of TaNGv1.1 for Genome-wide association studies (GWAS) and its utility for Copy Number Variation (CNV) analysis. The array is commercially available with supporting marker annotations and initial genotyping results freely available.

Keywords: Triticum aestivum; Axiom array; breeding; genotyping; single nucleotide polymorphism; wheat.

Figure 1

The percentage of probes in each probe quality category by array type. TaNG v1.1 has an increased ratio of the ‘high quality’ categories (Polymorphic High Resolution, No Minor Homozygous Group and Off‐Target Variants) and a decreased ratio of probes in the ‘low quality’ categories (Call Rate below Threshold, Monomorphic High Resolution and Other).

Figure 2

The physical distribution of chromosome 1 markers on the TaNG v1.1 (green) and 35K Wheat Breeder's Arrays (purple). (a) Number of markers in each of 20 bins spanning the chromosome. (b) Box and whisker plots of the number of markers per 10 Mb bin across the chromosome. There is a greater number of markers on the TaNG v1.1 Array (green boxes) and these are more evenly distributed than on the 35K Array (purple boxes). See Data S5 for plots of all chromosomes.

Figure 3

Comparison of Avalon × Cadenza genetic maps for chromosome 1A using (a) 35K Array map data and (b) TaNG v1.1 map data, showing the distribution and density of markers.

Figure 4

Comparison of chromosome location for markers based on physical and consensus assignment. Only markers with a consensus chromosome assignment are shown. Markers with physical positions derived from previous genotyping platforms (820K Array, 35K Array and DArT markers) are represented by bars. Markers derived from skim sequence data are represented by a line plot.

Figure 5

Copy Number Variance (CNV) frequency histogram for all samples genotyped with the TaNG v1.1 array in the initial screening (Data S4) across all chromosomes. Regions of copy number gain are displayed in the top track (blue) and regions of copy number loss are displayed on the bottom track (red) for each chromosome. Start and stop positions of each event are listed in Data S7.

Figure 6

Principle Component Analysis plot based on the USA material collections coloured by region of origin. Variation in PC1 and PC1 is 5.05 and 3.93, respectively. Samples used in breeding but of non‐USA origin were omitted from figure. Genotype data is available in Data S4.

Figure 7

Genome‐wide association study using the previous 35K Breeders Array, the new TaNG v1.1 array and the 10 million SNPs detected in sequenced data for three traits. (a) Heading date, (b) Leaf rust, (c) Stem rust. Identified QTL are highlighted in red.