Posters

Genes responsible for economic traits have been abundantly reported and still continue to be. For such known genes, breeding mostly aims to assess the allelic diversity captured in breeding collections and to identify individuals carrying favourable alleles at theses genes. Genotyping-by-sequencing (GBS) can identify and genotype thousands of single nucleotide polymorphisms (SNPs) on a genomewide scale, thus providing sufficient coverage of SNPs to define major haplotypes. Ultimately, these haplotypes can be equated to specific alleles of a given gene. Therefore, the identification of haplotypes from large SNP data sets represents a promising approach to routinely assess allelic variation in large collections. The definition of useful haplotypes is nonetheless challenging and relies on: i) the ability to define a relevant interval around the gene, ii) the ability to discriminate between informative variants (co-transmitted with the gene) from others (independently transmitted), iii) the ability to classify and illustrate resulting haplotypes for further interpretation. Our work aims to develop a computational approach able to extract and process haplotypes of targeted genes from dense SNP data. We are developing and calibrating our method on four well-characterized maturity genes. In this work, we demonstrate how our approach can provide accurate and essential information for breeding purpose by i) delivering a quick and clear picture of a gene’s allelic diversity in a given collection and, ii) accurately discerning groups of individuals sharing the same allele, iii) allowing to estimate an allele’s respective genetic background.