The invention discloses a grape titratable acid whole
genome selective breeding method. The method comprises the steps of
data acquisition and preprocessing, variation
site analysis,
whole genome association analysis and model analysis and prediction. The invention relates to the technical field of grape genetic breeding. Aiming at quantitative character characteristics (influenced by multiple genes and environment) of titratable acidic characters, more than 10000 significant variation sites and key
gene areas (such as candidate genes for coding
malic enzyme and
tartaric acid synthetase) covering
tartaric acid and
malic acid metabolism related pathways are positioned through
whole genome association analysis, so that the one-sidedness of
single gene analysis is avoided, and the quantitative character characteristics of the titratable acidic characters are obtained. The probability of false positive results is remarkably reduced, a model is constructed based on two-year
phenotype data, and the influence of
environmental noise is eliminated through
cross validation, so that a prediction result is closer to a real genetic effect, and the problem that traditional
phenotype selection is interfered by environmental fluctuation is solved.