A high-density integrated map for grapevine based on three mapping populations genotyped by the Vitis18K SNP chip

Abstract

Key message We present a high-density integrated map for grapevine, allowing refinement and improved understanding of the grapevine genome, while demonstrating the applicability of the Vitis18K SNP chip for linkage mapping. Abstract The improvement of grapevine through biotechnology requires identification of the molecular bases of target traits by studying marker-trait associations. The Vitis18K SNP chip provides a useful genotyping tool for genome-wide marker analysis. Most linkage maps are based on single mapping populations, but an integrated map can increase marker density and show order conservation. Here we present an integrated map based on three mapping populations. The parents consist of the well-known wine cultivars ‘Cabernet Sauvignon’, ‘Corvina’ and ‘Rhine Riesling’, the lesser-known wine variety ‘Deckrot’, and a table grape selection, G1-7720. Three high-density population maps with an average inter-locus gap ranging from 0.74 to 0.99 cM were developed. These maps show high correlations (0.9965–0.9971) with the reference assembly, containing only 93 markers with large order discrepancies compared to expected physical positions, of which a third is consistent across multiple populations. Moreover, the genetic data aid the further refinement of the grapevine genome assembly, by anchoring 104 yet unanchored scaffolds. From these population maps, an integrated map was constructed which includes 6697 molecular markers and reduces the inter-locus gap distance to 0.60 cM, resulting in the densest integrated map for grapevine thus far. A small number of discrepancies, mainly of short distance, involve 88 markers that remain conflictual across maps. The integrated map shows similar collinearity to the reference assembly (0.9974) as the single maps. This high-density map increases our understanding of the grapevine genome and provides a useful tool for its further characterization and the dissection of complex traits.

Grapevine genome analysis demonstrates the role of gene copy number variation in the formation of monoterpenes

“…The volatile profiles of the parent cultivars were compared ( Figure 1 ) and show clear segregation. Overall, the table grape selection, G1-7720, which is categorised as having a muscat aroma ( Vervalle et al., 2022 ), has a higher volatile content regardless of developmental stage. While, the terpene content of both parents is highest at veraison, G1-7720 has a significant increase in sesquiterpene content from pre-veraison to veraison berries.…”

“…A total of 137 progeny of the ‘Deckrot’ x G1-7720 mapping population have previously been genotyped with the Vitis18K SNP chip and 92 simple sequence repeat (SSR) markers ( Vervalle et al., 2022 ). The genotyping data were used to construct parental linkage maps for ‘Deckrot’ and G1-7720 with JoinMap ® 5 ( Van Ooijen, 2006 ).…”

“…This study utilises a biparental grapevine cross population established by crossing a wine cultivar and a table grape cultivar. A dense linkage map has previously been created for this mapping population ( Vervalle et al., 2022 ) which segregates for various traits, including aromatic profile. Quantification of volatile organic compounds in this population over several seasons was performed and genomic regions associated with these compounds identified.…”

Towards Marker-Assisted Breeding for Black Rot Bunch Resistance: Identification of a Major QTL in the Grapevine Cultivar ‘Merzling’

“…High-density genetic maps were generated with the software JoinMap ® 5.0 [ 31 ] by determining LGs through independence LOD. Map quality improvements were achieved following the good practices for datasets obtained by the GrapeReSeq Illumina 20K SNPchip described in Vervalle et al [ 95 ], i.e., chain length and stop criterion parameters optimization to allow convergence of the mapping algorithm. Firstly, an integrated map comprehending all robust SNPs and SSRs was calculated with Cross Pollination (CP) population type (wIN map).…”

“…Rather, the premature ungrouping—at lower LOD values compared to the other markers belonging to a linkage group—was used to exclude markers prior to grouping calculation. Other marker exclusion criteria applied to map construction were diagnostic values that stand out compared to the rest of the dataset, as high (i) nearest neighbor fit (NN Fit) and stress (NN Stress) that measure how well the markers fit that map order, and (ii) high -Log10(P) Genotype Probabilities Locus Means values, that measure how unlikely genotype observations affect the quality of the map [ 31 , 95 ]. Marker exclusion was checked through the function ‘combine maps’, to control if it corrupted the order of the map.…”