A High-Density SNP Genotyping Array for Genome-Based Breeding of Energy Sorghum for Central Europe

Sorghum is a promising alternative to maize for bioenergy production in Europe, however its use is still limited by poor adaptation to low temperatures during and after germination. Within the multinational European project ERANET-GAS (Genomic Adaptation of Sorghum) we are developing and applying genomic tools for bio-energy sorghum breeding. A major aim is utilization of genomic prediction strategies to identify and implement genetic variation associated with key adaptive and energy-related traits for European energy sorghum. Among the main target traits are early-stage chilling tolerance and dry matter yield. To establish a robust genotyping platform that delivers reproducible, high-quality SNP calls at loci spanning the genome at high density, we used next-generation sequencing data to stringently select high-quality, polymorphic SNPs for a new, 90k SNP Affymetrix Axiom genotyping array. The SNP selection was performed using around 265,000 SNPs from published GBS data, covering the diversity present within Sorghum bicolor, and from our own genomic resequencing data. Approximately equal numbers of genic and non-genic SNPs are represented, with priority given to array features with the highest predicted performance. 5,468 of the markers were assayed by two features, meaning that the array includes a total of 84,532 SNPs. We initially tested the array using 564 samples from heterozygous and homozygous sorghum breeding materials and segregating populations. After processing of the first sample plate using the Affymetrix power tool pipeline, a total of 73,569 SNPs (87%) passed stringent quality filtering criteria. Of these, 51,792 SNPs were subsequently scored as polymorphic high-resolution (PolyHighRes) markers. These provide a robust, low cost and rapid screening platform that is being applied for high-resolution linkage and LD mapping and used for prediction of hybrid performance in an ongoing breeding program.