A complication in the reuse of control groups or the inclusion of external controls arises when investigators wish to genotype cases on a platform different from the one used for controls. This may easily happen as genotyping technology changes and new chips with new pricing plans become available. It can appear necessary when funding is too limited to support a sufficiently powered study with both cases and controls genotyped together. Moreover, even if funding exists to genotype or re-genotype a control group on a particular chip, there may be limited biological samples available for use, or a desire to conserve such samples. However, while each platform genotypes a collection of tagging SNPs, different platforms choose these tagging SNPs in different ways. For example, Illumina uses patterns of linkage disequilibrium in the HapMap to choose its tagging SNPs, while Affymetrix (Affy) provides a large but less determinate collection of SNPs designed to give good coverage of the entire genome. There is not necessarily much overlap between the SNPs genotyped on two different platforms. For example, there were 140,325 SNPs in the
