Premium
Assessing genomewide statistical significance in linkage studies
Author(s) -
Lin D.Y.,
Zou Fei
Publication year - 2004
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/gepi.20017
Subject(s) - linkage (software) , statistic , computer science , nonparametric statistics , parametric statistics , genetic linkage , multivariate statistics , monte carlo method , missing data , statistical hypothesis testing , quantitative trait locus , data mining , computational biology , statistics , biology , genetics , mathematics , machine learning , gene
Assessment of genomewide statistical significance in multipoint linkage analysis is a thorny problem. The existing analytical solutions rely on strong assumptions (i.e., infinitely dense or equally spaced genetic markers that are fully informative and completely observed, and a single type of relative pair) which are rarely satisfied in real human studies, while simulation‐based methods are computationally intensive and may not be applicable to complex data structures and sophisticated genetic models. Here, we propose a conceptually simple and numerically efficient Monte Carlo procedure for determining genomewide significance levels that is applicable to all linkage studies. The pedigree structure is completely general; the marker data are totally arbitrary in respect to number, spacing, informativeness, and missingness; the trait can be qualitative, quantitative, or multivariate; the alternative hypothesis can be two‐sided or one‐sided; and the statistic can be parametric or nonparametric. The usefulness of the proposed approach is demonstrated through extensive simulation studies and an application to the nuclear family data from the Tenth Genetic Analysis Workshop. © 2004 Wiley‐Liss, Inc.