Causal inference frameworks in biology

I noticed that when doctors, economists, and computer scientists try to reason about the effects of genes, their thinking bottoms out in probabilities or effect sizes. Something like “this gene causes you to have 50% higher risk of cancer” or “these alleles will give you +5 IQ points”.

I feel uneasy seeing these kinds of claims, but it’s hard to say exactly why. These kinds of causal claims are very different from the ones we make in the lab. If I had to put my finger on it, I’d call them “predictive causality”, as opposed to “mechanistic