Figure 5.

Funnel plots of minor allele frequency corrected genetic associations with exposure (${\widehat{\gamma }}_j^C$) against causal estimates based on each genetic variant individually (${\widehat{\beta }}_j$) for 50 IV estimates in four scenarios: (a) no pleiotropy; (b) balanced pleiotropy; (c) directional pleiotropy, InSIDE assumption satisfied; and (d) directional pleiotropy, InSIDE assumption not satisfied. The inverse-variance weighted (IVW, red) and MR-Egger (blue) causal effect estimates are also shown.