Some more threads on human genetics. Firstly, as many people have noted, having a diverse group of scientists doing human genetics is healthy - it is healthy for *all of science* - just having diversity is good for team science

It is particularly health in human genetics because ultimately one has to part guess hypotheses about why things happen and then test (this goes to the very grungy aspect of "what covariates are we going to use to model human biology").

When one does observational studies on humans (human genetics, human epidemiology, human physiology) one immediately is forced to make modelling choices for human society (Socio-Economic criteria; Townsend indexes for UK; location of testing centre as covariates)

You can't dodge the issue - you can try to "model it out" to get at a cleaner genetics/biology readout - but even in the modelling one needs to model it well. Here one needs a diversity of scientists from a diversity of backgrounds to do the best job

(there is also the given, moral, side that one should be encouraging a diversity of scientists anyway and really science is for everyone. It should be a given, but it is often not).

The area of promoting diversity is more complex than it looks; lots of this is due to unconscious biases that basically everyone has (myself included!). The first thing is accepting that, and accepting that the sum of a lot of unconscious biases adds up to a pretty fearsome wall

The super annoying thing about unconscious bias is that it is ... unconscious and thus hard to spot yourself. This is both where processes is good (externalising issues to discuss and harness) and doing things in groups where different people spot different blind spots.

To stress, I am nowhere near free from bias. It irritates the hell out of me when I think about it, but I know the way is to think about process and have a diverse group of people around me.

Diverse groups of scientists I think also help lift the ambition and the role models for people around them. I'm super happy about the diverse groups of scientists I've worked with, both inside @embl and outside but its nowhere near the levels it should be. The road is long here

Second thing about human genetics is we still don't do nearly enough sub-Saharan human genetics. It is a real thing to push for. As most geneticists know the genetic diversity *and* recombination patterns are far more/deeper in sub saharan africa

Even if one took the v. narrow view that one only wanted to inform on "long haplotypes mainly found in European genetics" and somehow persuaded yourself that all these complex populations (black british in liverpool etc) didn't matter (which is madness!) it would still be ideal

Given that in fact humans are just one big complex mixture of genetics, with all sorts going on now and in fact always have been going in history on it is really a no-brainer not to do more sub saharan human genetics.

Re-echo'ing the point about, as soon as one does this one has to have diverse teams who understand human societies in these locations analysing the data - for equity reasons but far more importantly in the "knowing what one should model" reasons. Madness not to.

This is why I've always been so keen for @H3Africa and @H3ABioNet projects, and loved the whole embracing of human genetic diversity by other colleagues (eg, from @institutpasteur @QuintanaMurci's excellent work). It is super cool as well.

There is plenty other places to get better at human genetics - the Sahel communities, South America with all its complexity, and India stand out.

This all needs to be handled well; the right people formed in the right teams, right engagement, right language. Its complex, but it is what needs to be done.