Skip to main content
Fig. 7 | BMC Medical Research Methodology

Fig. 7

From: Overlapping-sample Mendelian randomisation with multiple exposures: a Bayesian approach

Fig. 7

Coverage from simulated data using our Bayesian method and classic methods (2SLS when sample overlapping rate = 1 and IVW otherwise) when β=(β1,β2,β3)=0.3 and δ=(δ1,δ2,δ3,δ4)=1. Results with two instrument strengths (top panels: α=(α1,α2,α3)=0.5 and bottom panels: α=0.1) are displayed, from left to right, for estimated causal effect of X1,X2 and X3 on Y, denoted as \(\hat {\beta }_{1}, \hat {\beta }_{2}\) and \(\hat {\beta }_{3}\) respectively. Each panel consists results of six different sample overlapping rates ranging from 0 (leftmost) to 1 (rightmost) and each box plot represents causal effect estimated from 200 simulated datasets

Back to article page