HEIDI-outlier. Nevertheless, these undetected pleiotropic effects do not seem to bias the GSMR estimate (Supplementary Fig. 4b), in contrast to a small bias in the estimate from Egger regression (MR-Egger) which is thought to be free of confounding from pleiotropy13. Our simulation results also show that the GSMR estimate of bxy is not significantly different from zero under a pleiotropic model without causal effect in the presence or absence of LD (Supplementary Table 2).
