Cleaning genotype data

The identification of genes contributing to variation in complex phenotypes requires genetic data of high fidelity. Thus, the identification of pedigr ee and genotyping errors is a crucial prerequisite to the analysis of data from a genome scan for disease genes. The problem has been given little att ention in most gene hunting papers; the focus has often been on eliminating mendelian inconsistencies in order that the analysis may proceed, rather t han on achieving the best possible data. Though a number of computer progra ms are available to assist in the identification of genotyping and pedigree errors, the process is still not completely automated. While the Collabora tive Study on the Genetics of Alcoholism (COGA) data set for GAW11 is compl etely compatible with Mendel's rules, there are still some errors present. We inspected the COGA data for the presence of additional errors, and ident ified five possible pedigree errors. (C) 1999 Wiley-Liss, Inc.