Data from: Genomics assisted ancestry deconvolution in grape

The genus Vitis (the grapevine) is a group of highly diverse, diploid
woody perennial vines consisting of approximately 60 species from across
the northern hemisphere. It is the world’s most valuable horticultural
crop with ~8 million hectares planted, most of which is processed into
wine. To gain insights into the use of wild Vitis species during the past
century of interspecific grape breeding and to provide a foundation for
marker-assisted breeding programmes, we present a principal components
analysis (PCA) based ancestry estimation method to calculate admixture
proportions of hybrid grapes in the United States Department of
Agriculture grape germplasm collection using genome-wide polymorphism
data. We find that grape breeders have backcrossed to both the
domesticated V. vinifera and wild Vitis species and that reasonably
accurate genome-wide ancestry estimation can be performed on interspecific
Vitis hybrids using a panel of fewer than 50 ancestry informative markers
(AIMs). We compare measures of ancestry informativeness used in selecting
SNP panels for two-way admixture estimation, and verify the accuracy of
our method on simulated populations of admixed offspring. Our method of
ancestry deconvolution provides a first step towards selection at the seed
or seedling stage for desirable admixture profiles, which will facilitate
marker-assisted breeding that aims to introgress traits from wild Vitis
species while retaining the desirable characteristics of elite V. vinifera
cultivars.