New approach to association testing in case-parent designs under informative parental missingness

The transmission/disequilibrium test (TDT) and related methods using genotype data on diseased probands and their both parents (triads) have been popular for testing linkage or association between a disease and a candidate gene. The usefulness of the TDT-type approaches lies mainly in their robustness, in the sense that they are valid under population stratification, arbitrary parental genotype distribution, and "informative" parental missingness where the parental missingness may depend on parental genotypes. Recently, a variety of extended TDTs were developed to accommodate parental missingness and to allow for using incomplete triads, including single-offspring-single-parent families (dyads) and single offspring with no parents (monads). However, these methods usually do not preserve the full robustness of the original TDT. In this paper, we propose a new TDT-type approach based on the conditional likelihood of the proband's genotype given the number and, if any, genotypes of the available parents, as well as the proband's phenotype. This new proposal keeps the full robust property of the original TDT. In addition, the new method is very easy to implement, without the need to specify models on parental mating-type probabilities and on parental missingness.