P-050 - Harnessing k-mers for Reference-Free Pangenome-Wide Association Mapping to Uncover Disease Resistance Genes in Wheat

The genetic diversity of commercial wheat varieties has been diminished by domestication and breeding, leaving them increasingly vulnerable to the growing stresses of climate change. In contrast, bread wheat landraces have higher genetic diversity likely to be crucial for adaptation. To exploit this untapped genetic diversity potential, we assembled and genotyped with DArTseq technology a diverse collection of 750 Spanish tetraploid wheat landraces to identify sources of resistance against the wheat yield-reducing fungal disease powdery mildew.

To move from the reliance on a single “reference” genome to define genetic variants, a k-mer-based pangenome-wide association mapping pipeline was developed. By developing a k-mer-based GWAS approach using fully assembled genomes of Triticum aestivum and its progenitors, we uncovered 25% more resistance-associated k-mers than single-reference methods, detecting some known Pm genes (e.g., Pm1, Pm3b, or Pm4b) but also multiple novel regions associated with powdery mildew resistance.

Of note, unlike standard GWAS, our method identifies associations with structural variations and sites not present in a single reference genome, outperforming SNP-based GWAS in loci detection and mapping precision. Further, it underscores the crucial role of landraces stored at genebanks as a rich source of novel genetic variation not only for mildew resistance but also for any other agronomic trait vital for adaptation to a changing environment.