As I got my vaccine boost yesterday I was talking to the nice clinician researcher (midwifery expert - interesting person) and we swapped research notes - in particular as we waited that I didn't have a severe adverse reaction in the first 15mins (always an interesting moment!)

He did implementation research - how best to get best practice implemented. A bit meta but super interesting. I sketched out the recent work on heart physiology genetics and up coming work on eye genetics from my group (I saved my love of Medaka fish for another time)

He was skeptical genetics could really tell us anything important about obs/gynae in particular in his practical end of the business, and I realised that many people thinking about genetics only being useful as an explanation not as a tool to understand biology

The basic explanation thesis is very much a proportion of variance explained - if genetics doesn't explain much variance of what you are looking at, why use genetics? First off of course there is much more variance explained in many phenotypes by genetics than people realise but>

<as importantly, the variance explained is really a ratio of effects (with this crappy genetics term "environment" which actually means "everything that is not genetics"), and it's sort of orthogonal to the insight genetics gives you on the biological processes behind something

Those processes might be molecular or cellular and increasingly tissue, organ and physiological (I refer you to this great recent paper I was a co-author on, lead by the brilliant @HannahVMeyer on heart physiology).

Here - and more generally - genetics is a tool to unpick the complexity of biology, and a really very powerful one

As nearly every molecule in our body has some relationship to our genome (direct or indirect), and nearly every part of genome varies (at some frequency+effect) genetics can be used as a very "hypothesis generating" scheme, encompassing much of biology

(to handle the low frequency of some variants and/or their low effect sizes one often needs large cohort sizes. This "throw large numbers of people" to counteract these issues feels sometimes inelegant, but ... it works robustly)

Furthermore as the assignment of alleles to parents to children is a very random process, we can leverage this randomness to have "probes" into biology which are largely unconfounded to all the other complexities of human existence. Mendelian Randomisation is one such approach.

(though there are a host of gotchas if you do this wrong; ideally get a battlescarred statistical geneticist or genetic epidemiologist on your team and then earn your own scars if doing this the first time)

Genetics is not magic pixie dust. It is definitely *not* your destiny (the modern day, pseudo-astrological phrase "its in our DNA" should be banished!) and although it can provide hypothesis-free insight, often that insight is "this bit of biology is freakin' complex"

However, genetics is a very useful tool. Used well it often gives key insights into any biological question which involves living organisms that have DNA (and ideally have simple diploid genetics; plant+fungi researchers can extend to the madness of plant+fungi funky genetics!)