As discussed, traditional case-control studies on GWA often include only unrelated individuals. By including family information in the study, we can expect an increase in power for linkage and association detection. On the other hand, because the traditional TDT requires complete genotypic information from a trio by measuring over-transmission of an allele from heterozygous parents to the affected offspring, it may be less useful in trio data like COGA where there are relatively few complete trios. To overcome these limitations, in this project, we extend the original TDT to the dTDT to accommodate dosage probabilities of a trio. The trio-dosage sets can be inferred through programs like MERLIN. Compared to a recent work from Edenberg et al, the dTDT shows increased power to detect association.
