Grapes (Vitis vinifera subsp. vinifera) are among the most important horticultural crops in the world. The Vitis genus is an ideal model system to study marker design for the following reasons: (1) The genetic background is highly diverse in the genus Vitis. This genus, which contains approximately sixty species, diverged around 28 million years ago. It is widely distributed in Asia, Europe, North and South America. These locally adapted wild species provide breeders with an abundant resource of germplasm to improve biotic and abiotic tolerance and resistance in cultivated grapes. To make use of wild species, molecular markers that can be used across diverse wild species are essential to the breeding practice. (2) Grapes are a highly heterozygous diploid species. Many economically important plant are highly heterozygous species, including apples\cite{Bianco2016}, pears \cite{wu2013genome}, raspberries \cite{Graham2004}, bananas, sweet oranges \cite{Jiao2013}, Cassava\cite{Perea2016}, wild potatoes \cite{Leisner2018}, sweet potatoes\cite{Si2016}, and many woody species . Genotyping for highly heterozygous species is challenging in general due to their high genetic diversity and fast linkage disequilibrium decay. The lessons we learn in the Vitis genus can be applied to these highly heterozygous species. (3) In crops, the grape genome size is comparatively small (maize is 2 Gbp, while grape is 450 Mbp).
Much progress has been made to address the difficulties in genotyping highly diverse and heterozygous species using high-through put sequencing. Resequencing the entire genome is too costly and the marker density exceeds the needs of genetic studies thousand of times. To decrease the genome complexity, restriction enzymes are used to restrict the sequencing in regions that can be digested by the restriction enzymes. However, the missing rate and XX is high in GBS. One solution has been proposed to further enrich the target region by a capture array \cite{Uitdewilligen2013} .
For example, one approach is that enriching the target region by a capture array and then conduct \cite{Uitdewilligen2013}genotyping-by-sequencing .