The use of covariates in affected sib pair analyses

Research output: Contribution to journalArticlepeer-review

Abstract

Linkage analysis and association studies are two basic strategies currently used to detect and map susceptibility genes for complex traits. Recent linkage methods have relied on haplotype sharing within affected sibling pairs (or sets of affected relatives). Three areas of concern are (1) how to handle phenotype definition - multiple parallel analyses are typically performed; (2) how to combine multiple data sets in a meta-analysis - current individual data sets may lack power; and (3) how to perform conditional analysis where sharing on one chromosome influences sharing at another. One method to address these problems is to assume that the probability of a sib pair sharing an allele from a particular parent has a logistic regression on a set of covariates. In this setting, covariates may include indicators for diagnosis, indicators for source of data in a meta-analysis, or genotypes at a particular locus. This provides a general setting for an overall test of linkage, and individual tests of phenotypic and study heterogeneity. The use of multipoint marker data allows the estimation of gene location. We describe general sib pairs analysis, the extension to allow for covariates, and then illustrate these methods with data from COGA, the Collaboration Study on the Genetics of Alcoholism.

Original languageEnglish
Pages (from-to)460-461
Number of pages2
JournalAmerican Journal of Medical Genetics - Neuropsychiatric Genetics
Volume81
Issue number6
StatePublished - Nov 6 1998

Fingerprint

Dive into the research topics of 'The use of covariates in affected sib pair analyses'. Together they form a unique fingerprint.

Cite this