Transmission disequilibrium test with discordant sib pairs when parents are available

Abstract.

The transmission disequilibrium test (TDT) has been employed to map disease susceptibility loci (DSL), while being immune to the problem of population admixture. The customary TDT test (TDT_D) was developed for affected child(ren) and their parents and was most often applied to case-parent trios. Recently, the TDT has been extended to the situations when (1) parents are not available but affected and nonaffected sibs from each family are available, (2) unrelated control-parent trios are available for combined analyses with case-parent trios (TDT_DC), and (3) large pedigrees. For many diseases, affected children in the case-parent trios enlisted into the TDT_D have unaffected sibs who can be recruited. We present an extension of the TDT by effectively incorporating one unaffected sib of each of the affected children in the case-parent trios into a single analysis (TDT_DS, where DS denotes discordant sib pairs). We have developed a general analytical method for computing the statistical power of the TDT_DS under any genetic model, the accuracy of which is validated by computer simulations. We compare the power of the TDT_D, TDT_DC, and TDT_DS under a range of parameter space and genetic models. We find that the TDT_DS is generally more powerful than the TDT_DC and TDT_D, particularly when the disease is prevalent (>30%) in the population. The relative power of the TDT_D and the TDT_DS largely depends upon the allele frequencies and genetic effects at the DSL, whereas the recombination rate, the degree of linkage disequilibrium, and the marker allele frequencies have little effect. Importantly, the TDT_DS not only may be more powerful, it also has the advantage of being able to test for segregation distortion that may yield false linkage/association in the TDT_D.