Finally, it is important to note that GWAS is but one use of genome wide data. For instance, the genome-wide relationship models discussed above (Yang et al., 2010) can be extended to incorporate these multilevel design features so that the contextual variation in the heritability of health behaviors can be examined using measured genetic similarity rather than assumed similarity from sibling-based models. Similarly, sibling fixed effects approaches can take advantage of the “random assignment” of risk alleles to examine siblings residing and socializing in similar environments compared to those in very different social contexts (Fletcher et al., 2011). These methods provide unique and new possibilities to identify causal models and have thus far not been extended to the genome-wide level.
