Fig. 2: Identifiability of the direct and indirect causal effects between traits under three scenarios.
From: Causal mediation analysis for time-varying heritable risk factors with Mendelian randomization
a Only X1 has valid instruments and the direct and indirect effects are not identifiable. b All exposures have valid instruments, and both the direct and indirect effects are identifiable. c None of the exposures have valid instruments, but the effects remain identifiable under an independence assumption.
