One obvious limitation of what we have proposed so far is that the genetic variants related to the biological intermediate need to be known a priori in order for the approach to work. In addition, in the case of intermediates where known variants exist, they may explain only a small amount of the total phenotypic variance in that variable. However, we and others have previously shown that genome-wide allelic scores generated by simply counting up hundreds of thousands of anonymous “risk” alleles in genome-wide SNP data are capable of explaining meaningful amounts of phenotypic variance in traits of interest [9], [10]. Our idea is to use these genome-wide allelic scores in situations where there are no known confirmed genetic variants and/or in situations where the known variants explain inadequate proportions of the phenotypic variance in the biological intermediates of interest. In fact, our previous work has shown that these scores can explain more phenotypic variance than allelic scores constructed from confirmed variants only [9], [10]. This is because many complex phenotypes (including biological intermediates) are influenced by hundreds, if not
