Pooling data and linkage analysis in the chromosome 5q candidate region for asthma

We investigated a variety of methods for pooling data from eight data sets (n = 5,424 subjects) to validate evidence for linkage of markers in the cyt okine cluster on chromosome 5q31-33 to asthma and asthma-associated phenoty pes. Chromosome 5 markers were integrated into current genetic linkage and physical maps, and a consensus map was constructed to facilitate effective data pooling. To provide more informative phenotypes with better distributi onal properties, variance component models were fitted using Gibbs sampling methods in order to generate residual additive genetic effects, or sigma-s quared-A-random-effects (SSARs), which were used as derived phenotypes in s ubsequent linkage analyses. Multipoint estimates of alleles shared identica lly by descent (IBD) were computed for all full sibling pairs. Linkage anal yses were performed with a new Haseman-Elston method that uses generalized- least-squares and a weighted combination of the mean-corrected trait-sum sq uared and trait-difference squared as the dependent variable. Analyses were performed with all data sets pooled together, and also separately with the resulting linkage statistics pooled by several meta-analytic methods. Our results provide no significant evidence that loci conferring susceptibility to asthma affection or atopy, as measured by total serum IgE levels, are p resent in the 5q31-33 region. This study has provided a clearer understandi ng of the significance, or lack of significance, of the 5q31-33 region in a sthma genetics for the phenotypes studied. (C) 2001 Wiley-Liss, Inc.