All SNPs can be invalid due to pleiotropy as long as the pleiotropy satisfies the InSIDE assumption. Can be implemented in a Bayesian framework. Standard implementation requires gene-exposure associations to be orientated in positive direction. Radial MR formulation avoids this step. The intercept of a MR-Egger regression provides an indication of pleiotropy when it is not null. Similar to IVW methods, the NOME assumption still holds and, if violated, there may be weak instrument bias. This can be tested with the I2 statistic with the MR-Egger method.
- Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. International Journal of Epidemiology 2015;44:512-525.
- Bowden J, Spiller W, Del Greco M F et al. Improving the visualization, interpretation and analysis of two- sample summary data Mendelian randomization via the Radial plot and Radial regression. International Journal of Epidemiology 2018;47:1264-1278.
- Schmidt AF, Dudbridge F. Mendelian randomization with Egger pleiotropy correction and weakly informative Bayesian priors. International Journal of Epidemiology 2018;47:1217-1228.
- Bowden J, Del Greco MF, Minelli C, Davey Smith G, Sheehan NA, Thompson JR. Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic. Int J Epidemiol 2016:doi:10.1093/ije/dyw220.