and control individuals studied in the present sample were collected at the same site, recruited in ways that result in virtually all candidates consenting to participation, and assessed using the same instruments. Controls are thus characterized in such a manner that they each provide a contrast with the dependent cases. 2a) Genotyping data in dbGAP typically uses data from single microarrays that are hybridized with fluorescently-labeled DNAs prepared from DNAs from single individuals, hybridization intensities assessed, and genotype calls made based on Bayesian and other algorithms using data from the ratios of hybridization intensities to probes that are complementary to alternative allelic forms of each SNP. Quality control efforts for samples and SNPs use predetermined algorithms for hybridization signal differences and analyses of Hardy-Weinberg equilibria. Few studies provide test-retest data to evaluate the fraction of genotypes that are replicable, however. 2b) The current approach uses data from three microarrays that are hybridized with three distinct preparations of fluorescently labeled DNA that are carefully prepared from pools of DNA from 20 individuals. Quality control evaluations come chiefly from assessment of the array-to-array variation in hybridization intensities noted for replicate experiments, as well as assessments of pool-to-pool variation, as noted here. These
