Premium
Robust LOD scores for variance component‐based linkage analysis
Author(s) -
Blangero John,
Williams Jeff T.,
Almasy Laura
Publication year - 2000
Publication title -
genetic epidemiology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.301
H-Index - 98
eISSN - 1098-2272
pISSN - 0741-0395
DOI - 10.1002/1098-2272(2000)19:1+<::aid-gepi2>3.0.co;2-y
Subject(s) - kurtosis , statistics , linkage (software) , mathematics , normality , quantitative trait locus , variance (accounting) , multivariate normal distribution , multivariate statistics , heritability , pedigree chart , trait , skewness , econometrics , type i and type ii errors , computer science , biology , genetics , accounting , gene , business , programming language
The variance component method is now widely used for linkage analysis of quantitative traits. Although this approach offers many advantages, the importance of the underlying assumption of multivariate normality of the trait distribution within pedigrees has not been studied extensively. Simulation studies have shown that traits with leptokurtic distributions yield linkage test statistics that exhibit excessive Type I error when analyzed naively. We derive analytical formulae relating the deviation from the expected asymptotic distribution of the lod score to the kurtosis and total heritability of the quantitative trait. A simple correction constant yields a robust lod score for any deviation from normality and for any pedigree structure, and effectively eliminates the problem of inflated Type I error due to misspecification of the underlying probability model in variance component‐based linkage analysis. Genet. Epidemiol. 19(Suppl 1):S8–S14, 2000. © 2000 Wiley‐Liss, Inc.