SNP causality in a region of Q variants can be summarised for each trait using a vector of length Q of (0, 1) values, where 1 means that the variant is causally associated with the trait of interest and at most one entry is non-zero. A schematic illustration of this framework is provided in Figure 1 in a region that contains 8 SNPs. Each possible pair of vectors (for traits 1 and 2, which we refer to as “configuration”) can be assigned to one of five hypotheses:
