Some of the methods included in our study have been compared to one another before. MV-PLINK and MultiPhen were compared by O'Reilly et al. and van der Sluis et al., which showed that both methods result in the same power when restricting the analysis to normally distributed traits [12], [15], corroborating our findings. TATES was compared to MANOVA (which is similar to MV-PLINK) and shown to be only outperformed by MANOVA in the particular condition that the genetic variant affects only one of multiple strongly correlated traits [15]. In contrast, we observed that MV-PLINK outperformed TATES in almost all scenarios.
