Two versions of the data were provided to GAW (Tables 1 and 2): raw data and clean data with Mendelian inconsistencies removed. Genotyping data was ordered using the maps provided by the companies and merged with the COGA family data to make the raw comma delimited data files. PEDCHECK [34] was used to detect Mendelian inconsistencies for each SNP. Level 0 and 1 checks were run, and Mendelian inconsistencies were removed in each nuclear family according to the rules listed below:
