Potato improvement with markers: the value of read-backed haplotypes

This thesis investigates the use of low-cost, low-density amplicon-based genotyping to support genomic applications in potato breeding. We developed PotatoMASH, a platform which simultaneously generates SNPs and read-backed haplotypes (haplotags) at 339 loci and demonstrated its use for QTL discovery, heterozygosity tracking, and genomic prediction by genotyping over 1800 clones from several breeding programmes. A core question was how haplotags perform relative to their constituent SNPs. In a diploid breeding panel, haplotags complemented SNPs in QTL detection, with both marker types together identifying 37 QTL for 23 traits. Haplotags also provided more accurate homozygosity estimates than SNPs, supporting Fixation-Restitution breeding strategy that combine allele fixation with the preservation of strategic heterozygosity. PotatoMASH delivered reasonable genomic prediction accuracy at a fraction of the cost of high-density platforms. By unifying GWAS, MAS, inbreeding analysis, and genomic selection within a single system, this work demonstrates that read-backed haplotypes are a powerful tool to advance breeding-by-design in potato.