| 摘要 |
The present invention provides a trait prediction model creation method for predicting the phenotype of a trait from single nucleotide polymorphism data, and a trait prediction method for enabling prediction of the trait at high probability. The trait prediction model creation method comprises: a step for representing, in a matrix, a plurality of single nucleotide polymorphisms by using the single nucleotide polymorphism data in which a plurality of single nucleotide polymorphisms and a plurality of traits are associated with each other in a plurality of organisms; a step for classifying the plurality of single nucleotide polymorphisms into a plurality of categories on the basis of the genetic structures; a step for calculating, for each of the categories, a genomic similarity matrix by using the matrix representation and the number of single nucleotide polymorphisms belonging to the category; and a step for applying the genomic similarity matrix and parameters of the genetic structures to a linear mixture model. |
