and to include both related and unrelated subjects [76–78]. The FBAT generalizes the TDT by computing the covariance between genotype and phenotype that are centered in such a way to accommodate different sampling designs. The method can be used to study survival traits [79] in addition to qualitative and quantitative phenotypes [80] and general pedigrees. The FBAT is robust to population stratification because of the use of controls within family members who share the same genetic background. Extensions of the TDT that include the founder genotypes to increase the power of the study have been proposed but do not protect against population stratification [81]. The FBAT does not require any parametric assumption about the phenotype, and the null hypothesis of no association can be tested using large sample approximations that appear to work well with at least 10 informative families [50]. Although this model-free feature makes the test robust to model misspecifications, it limits its use to test for associations that can be at most adjusted for the effect of other covariates such as gender or environmental effects, but neither gene–gene interactions nor gene-environment interactions can be tested. Model-based approaches overcome these limitations by directly modeling the effect of the
