We used two GWAS datasets for the same protein, one GWAS from the UKB-PPP, specifically the European (dsicovery), and the other from the deCODE. Surprisingly, the effect estimates derived from a mendelian randomization analysis were in opposite direction, one increases the risk of a certain disease and the other decreases the risk. I wanted to know if this issue has been raised by any other scientists who used GWAS from two different sources, and if there is any explanation/justification to this finding. Thank you

Created by Zoheir Alayash ZoheirAlayash
Please contact the corresponding authors of the Nature publication for feedback on downstream analyses.

Applying mendelian randomization analysis using the UKB GWAS we get contradictory results when we used GWAS from the deCODE page is loading…