Linkage and association analyses of alcoholism using a regression-based transmission/disequilibrium test

Recently, George et al. proposed a regression-based transmission/disequilib rium test for linkage using information on the parent-to-offspring transmis sion status of an allele at a marker locus. We extended this test by simult aneously testing for any population association by incorporating the presen ce/absence status of the associated allele as a covariate in the model. We used this method to analyze markers on chromosomes 1 through 21 of the Coll aborative Study on the Genetics of Alcoholism data on alcoholism for possib le association and linkage. We found nominal significance (at the 0.02 leve l) at eight different regions for linkage, though statistical significance may not be concluded due to multiple testing. The strongest evidence of lin kage was observed for markers D4S2639 and D12S397 with p-values less than 0 .005. We also found strong association between the trait and alleles 149 of D7S691 and 131 of D21S1437. (C) 1999 Wiley-Liss, Inc.