Genomic Profiling of Plum Prunus domestica L. Germplasm using Genotyping-by-Sequencing

European plum P. domestica L. (2n=6x=48) is a commercially important fruit tree species that is cultivated worldwide. With the rise of domestic and international market competition, knowledge on accurate varietal identities can be beneficial for plum breeders, producers and traders.  Here we attempted to establish a reliable discrimination system for plum varieties using genotyping-by-sequencing (GBS), a rapid and robust approach for reduced-representation sequencing of multiplexed samples. Plant material collected for this study includes 333 cultivars and landraces from the plum repository at INRA (Bordeaux, France), and 127 cultivars and selections held in the U.S.  In a pilot experiment, we genotyped 90 accessions representing different types of the European plum and its possible progenitors, diploid species P. spinosa L. and P. cerasifera L.  In order to capture appropriate sequencing depth while still affording a cost-effective multiplex design, we selected the Pst1 enzyme based ‘in silico’ digestion of the reference Prunus persica v 1.0 genome resulting in moderate cut frequency.  Unidirectional sequencing (SE-101 bp) of 90 pooled samples was carried out with an Illumina HiSeq 2500 across two flow-cell lanes, generating over 256M raw reads. Using in-house bioinformatic workflows compiled from public sources, we identified a subset of high-quality SNPs capable of varietal resolution in a polyploid species. Intermediate results of this study indicate robustness of the GBS approach for genomic profiling in plum and a need for introducing statistical parameters that allow discrimination between seedlings and vegetatively propagated clones in the background of a significant amount of missing data.