[gov3009-l] Sorensen on "Modelling components of variation"

Hi there, We hope you will join us this Wednesday, April 28th at the Applied Statistics workshop when we will be happy to have Daniel Sorensen from University of Aarhus. Details and an abstract are below. A light lunch will be served. Thanks! "Modelling components of variation: is environmental variance of quantitative traits genetically controlled?" Daniel Sorensen University of Aarhus April 28th, 2010, 12 noon K354 CGIS Knafel (1737 Cambridge St) Abstract: The classical model of quantitative genetics assumes that genotypes affect phenotypic values and means but that the variance of phenotype, given genotype (environmental variance) is the same for all genotypes. An extension (HET) postulates that both mean and variability differ between genotypes. Over recent years, statistical support for a genetic component at the level of the environmental variance has come from fitting the HET model to field or experimental data in various species. Since the skewness of the marginal distribution of the data under the HET model is directly proportional to the coefficient of correlation between genes affecting mean and variance, there is the concern that statistical support for the HET model may be an artifact of the scale of measurement. One may pose the question: Is there still support for the HET model when the data are analyzed in the “correct”scale? Here this was investigated by extending the HET model to accommodate the family of Box-Cox transformations. Litter size data in rabbits and pigs that had previously been analyzed in the untransformed scale were reanalyzed in a scale equal to the mode of the marginal posterior distribution of the Box-Cox parameter. In the rabbit data, the statistical evidence for a genetic component at the level of the environmental variance is considerably weaker than that resulting from an analysis in the original metric. In the pig data, the statistical evidence is stronger, but the coefficient of correlation between additive genetic effects affecting mean and variance changes sign, compared to the results in the untransformed scale. The study confirms that inferences on variances can be strongly affected by the presence of asymmetry in the distribution of the data. We recommend that to avoid one important source of spurious inferences, future work seeking support for a genetic component acting on environmental variation using a parametric approach based on normality assumptions, confirms that these are met. Cheers, matt.