Heatmap from Scientific Research

Adjacency matrix describing the Markov properties of the DAG involving the genetically predicted exposures and outcomes (the variables in the x axis are dependent on the variables in the y axis) that are function of the IVs and the inverse-variance weighting (IVW) (depicted with an asterisk) summary-level statistics B=(β,β,β)and B=(β,β,β). Neither reverse causation (top-right submatrix) nor phenotypic traits feedback loops (main diagonal) are allowed. Color code: black, directed edge between variables; white, no causal relationship between variables; black-white strips, directed edge not allowed (feedback loop and reverse causation between exposures and outcomes).

DOI