Analyses of the COGA data set in one ethnic group with examinations of alternative definitions of alcoholism

We used GENEHUNTER and GENEHUNTER-PLUS to search for linkage with the marke rs in the Collaborative Study on the Genetics of Alcoholism (COGA) data set in a single ethnic group. Analyses of a complex disorder such as alcoholis m depend on the definition of affection status. The COGA study provides two definitions of alcoholism (variables ALDX1 and ALDX2). To identify more se verely affected alcoholics that might be more homogeneous genetically, we d eveloped two other ways of characterizing subjects as alcohol dependent: (1 ) by combining the symptom variable values equally into a 24-point scale an d (2) by weighting optimally the symptoms and other descriptive variables i nto a single score using logistic regression. We applied these definitions within a single ethnic group to map alcoholism-related loci. We found two r egions on chromosome 1 that have adjacent markers significant at p-values l ess than or equal to 0.05. ALDX1 provided the highest Z-scores compared to the alternatives. (C) 1999 Wiley-Liss, Inc.