Code must be run in alphabetical order. For example : 1a_, then 1b_, then 2a_* The code is not self contained, because data bust maximum memory usage on GitHub. Hence, no data is given. However, link to the data can be found in the manuscript.

Code was performed in the R V.4.1.3 computing environment using publicly accessible functions from the TwoSampleMR V.0.5.6 https://github.com/MRCIEU/TwoSampleMR, The MendelianRandomization V.0.5.1 https://cran.rproject.org/web/packages/MendelianRandomization/index.html the data.table V.1.14.0 https://github.com/Rdatatable/data.table packages. The coloc package V.5.1.0 https://github.com/chr1swallace/coloc. The hyprcoloc package V.1.0 https://github.com/jrs95/hyprcoloc. The tidyverse V.1.3.1 collection of R packages, the gwasglue package V.0.0.0.9000 and the gwasvcf V.0.1.0 package was also used for data wrangling.