The TATES method is described in detail in the Materials and Methods section. Briefly, TATES requires the m×n p-values of the regression of m phenotypic variables on n GVs, and the m×m correlation matrix of the phenotypes. The regression of the phenotypes on the GVs can be conducted in standard software packages like PLINK, Mach2dat/qtl, SNPtest, and Gen/ProbABEL [13], [16]–[20], which are fast, facilitate quality control, and can correct for population stratification. For samples that include related individuals, analyses could be conducted using PLINK (where the –mperm option should not vary over the m phenotypes to assure that the p-values used in TATES have similar accuracy), *ABLE, PBAT or Merlin-offline [13], [16]–[17], [21]–[22]. For each GV, TATES sorts the m p-values ascendingly. To derive from these m p-values one trait-based p-value PT for each of the n GVs, TATES takes into account that the m phenotypes, and thus the m p-values, are correlated. In an iterative procedure, TATES weighs the j th p-value in the 1 to m sorted p-values with m e/m ej, where m e is the effective
