The methods to estimate the association can be categorized into two groups based on whether the trait is continuous or categorical. A continuous trait following a normal distribution is modeled as a function of the genetic effect using linear regression. General genotype association is tested using an analysis of variance that compares the distribution of the trait in the three genotypes MM, Mm, and mm [65]. A popular alternative is to represent the three genotypes by the variable X taking values 0 = MM, 1 = Mm, and 2 = mm and use this variable in linear regression. The parameterization is known as the additive genetic model, and the regression coefficient of the variable X represents the average change in the trait for each extra copy of the allele m, with a significant regression coefficient denoting a significant association. The additive model is easy to interpret and therefore the most commonly used. However, note that a lack of association in the additive model does not imply no association between the SNP and the trait. The genetic association can be adjusted
