Mendelian randomization (MR) has been criticized due to a flood of papers based on the output of a single R command.
But MR can offer valuable insights into disease etiology and can be really fun
Some interesting MR applications in examples from our work over the years 🧵 1/17
But MR can offer valuable insights into disease etiology and can be really fun
Some interesting MR applications in examples from our work over the years 🧵 1/17
Comments
Focusing on variation in a gene locus, MR can offer insights into the effects of genetic perturbation of a drug target
Next to the study of rare loss-of-function variants, this is probably the best way to provide genetic support for a drug target🧬
2/17
👉IL6R in cardiovascular disease
👉PCSK9 in stroke
👉BP-lowering targets in stroke subtypes
👉cytokines in inflammatory disease
There are several published examples in this space.
3/17
👉https://ahajournals.org/doi/full/10.1161/CIRCGEN.119.002872
👉https://atherosclerosis-journal.com/article/S0021-9150(22)01433-2/fulltext
👉https://neurology.org/doi/10.1212/WNL.0000000000009814
👉https://nature.com/articles/s42003-025-07453-w
4/17
As physicians are blinded to the genetic profile of individuals when prescribing a drug, a randomization takes place based on genetic proxies of response. This can be used to examine causal effects of a drug on side-effects or outcomes for repurposing
5/17
🔗https://academic.oup.com/brain/article/145/8/2677/6590424
6/17
MR can be used to test associations of a risk factor with an outcome across the spectrum of the risk factor.
Applications here:
👉effect of blood pressure on stroke risk
👉effect of IL6 signaling downregulation on cardiovascular risk across CRP values
7/17