Use of robust variance components models to analyse triglyceride data in families

Abstract

A robust approach for analysis of variance components models is presented which does not rely on the assumption of multivariate normality for its validity. This approach uses the multivariated normal distribution as a 'working model' but obtains standard errors for the final estimators which do not depend on this underlying distribution. By using the observed variance in the first derivatives of the multivariate normal 'working model' to modify the conventional score test, hypotheses regarding specific components can also be tested without relying directly on the assumption of multivariate normality. A special case is presented where both the modified score test and the likelihood ratio test are equally robust, and simulated data are used to illustrate this situation. Measurements of triglyceride levels in 391 individuals in 60 families randomly selected from the membership of a health maintenance organization are used to illustrate this robust approach to variance components.

Genetic epidemiology of fecal egg excretion during Schistosoma mansoni infection in an endemic area in Minas Gerais, Brazil

“…We used the pedigree based maximum likelihood variance decomposition method as implemented in the computer program SOLAR (Almasy 1998) to assess the relative importance of genetic factors in the determination of the variation of fecal egg counts. By assuming multivariate normality as a working model within pedigrees, the likelihood of any pedigree can be easily specified and numerical procedures can be used to estimate the variance component parameters (Beaty et al 1985). Likelihood estimation assuming multivariate normality yields consistent parameter estimates, even when the distributional assumptions are violated.…”

Genetic Determinants of Hemostasis Phenotypes in Spanish Families

“…Although we assume multivariate normality, this assumption is robust, and consistent parameter estimates are obtained when the assumption is violated. 15 Because 12 pedigrees were ascertained through a thrombophilic proband, we performed an ascertainment correction to obtain unbiased parameter estimates relevant to the general population. This was achieved by conditioning on the probands' phenotype.…”

Additive host genetic factors influence fecal egg excretion rates during Schistosoma mansoni infection in a rural area in Brazil.

“…We also used a statistical method (described below) that is robust to distributional violations when testing the nullity of single effects. 23 Pedigree identification. Evidence for the action of genes on a given trait derives from the covariances (or correlations) of the trait lnepg between different classes of relatives.…”