Did one my crash courses in human genetics + history today which is always fun, and tackled the "our intuition is bad about ethnic groups" - in other words, when some people hear things about "genetics" they often unconsciously elide them into "between ethnic group" differences
My starting point in breaking this down are the classic admixed populations - Cape Coloured (Cape of Good Hope, Portugal/Khosia/Bantu mix), Brazilian (Native American/European/West African), AfroCarribean (West African/European/a little Carrib)
Notice already this terminology is historical and biased - Afro Caribbeans should be "Afro-Euro-Carribbeans" (assuming the Caribbean is about geography and the other terms are about ancestry). Clearly these words are just the historical labels (its not principled genetics!)
This sort of story repeats around the world; in the UK the "Liverpool Black Community" - earlier African/European mixing due to the docks in Liverpool, and the study of Y chromosomes in British-identified Yorkshiremen which has some Y chromosomes of recent West African origin
(This starts to get people to realise that genetic mixing happens in both directions; perhaps not equally, but very rarely - I don't know of any examples - unidirectionally)
Indeed, even the genetic concept of a "European" is actually a mixture, just an older mixture - at least 3 different groups coming into Europe at different times and then eventually mixing reasonably evenly (though... the mixing process not perfect... red hair has a cline)
So - human populations, (basically) nearly always have sex and mix genetically *every time* they meet and both the current *and* the ancient world is full of movements of people and mixing
Back now to traits - we can measure all sorts of traits on humans - hair colour, height, weight, running speed, cholestrol levels, do you have schizophrenia, whether you stayed in education to college etc etc. Each of these have contributions from genetics
But the traits we use to classify (both self-classify and classify others) into "ethnic groups" - mainly hair colour and skin colour - both occupy small amounts of the genome.
So - each of these "identifying" traits *are* genetic, but their presence tells us far less than you expect about the other parts of the genome.
A dark skinned Bajan (from Barbados) man almost certainly has man recent african segments across his skin pigmentation loci, but it tells us only a little bit about the odds of being african at his other loci, and certainly a very small amount compared to a whiter skinned Bajan.
And we are *all* like this; we are all complex mixtures with complex histories. Some of the mixing has happened recently, some near recent, some old.
And the one thing we can be pretty sure about is that the differences in other traits we ascribe to ethnic groups which is *not* about skin colour+hair is extremely *unlikely* to be genetic, because we know many of these groups are in a "mid mixing" state
Now - I appreciate on twitter this sort of thread is mainly going to trigger (at some point) the charming 'race realists' (people who think there is a scientific basis to the ethnic groups). Apologies in advance for muting you if you simply rage against this.
It's not that I don't want to engage (otherwise why would I tweet this?) - its more that this thread is for people with the nagging "but wait a second, surely there must be *something* genetic about ethnic groups" but want to understand this more.
(For what it is worth; I am genuinely a genomics expert and I know the sub field that is really making progress in this area - human genetics and population genetics reasonably well, though there are many many people better than me in both of these sub-fields)
