My lab read this paper for journal club, and had some thoughts on the strong claims made about the number of signals of selection found. 1/biorxiv.org/content/10.110… The GLMM test that is applied in the paper, which gives the Z statistic and estimated selection coefficients, is looking for SNPs that show a consistent temporal allele frequency change. That is equivalent to Gene Environ correlation methods using time as the environment. 2/

The use of the phrase “ancestry-specific variant” is increasing, particularly to describe rare Single-nucleotide variants (SNVs). But these alleles are not ancestry-specific. They have not yet been found elsewhere, but they will be. 1/n The world has ~8 billion people. The mutation rate is ~10^-8 per base pair, so every base pair has mutated ~160 times in the past generation alone. Thus, every single (non-lethal) base pair mutation will be present in every large human grouping. 2/n

The All of Us paper is rightly being criticized for its UMAP figure, which suggests an overly discrete view of human variation—a problem that is compounded by colouring the plot with self-identified race and then omitting the “self-identified” from the title & legend. 1/n The paper presents a major NIH resource, but it does not take on board the carefully thought-through advice of the NAS panel that the NIH commissioned (presumably for exactly this kind of purpose). 2/n
nationalacademies.org/our-work/use-o…

The NAS report on the use of Population Descriptors in Human Genetics & Genomics is definitely worth a read by everyone working in human genetics and adjacent fields (chapter 5 is a good place to dive in if you’re short on time). 1/n
nap.nationalacademies.org/catalog/26902/… The report has a range of well argued recommendations, which are practically motivated & that we all can start to explore implementing in our research and papers. 2/n

The genetic ancestry labels that the NIH “All of Us” biobank genetic data returns seem likely to add to general confusion about genetics and race/ethnicity [see appendix A here: ] 1/nresearchallofus.org/wp-content/the… “All of Us” assigns study participants to one of a few discrete ancestry groups using their genome data (based on genetic similarity to the 1000 genomes & HGDP categories). 2/n

@kathleenmkay @DavidBLowry So if you're wanting to show a single locus, dominance variance is the easiest to add-in. @kathleenmkay @DavidBLowry You can say additive variance is the variance explained by assuming a linear relationship between genotype and phenotype (red line) and dominance variation is extra variance explained by the deviation away from additivity (dev. away from red line)

There appears to be some confusion generated by having this conversation on Twitter. There are, needless to say, a number of possible uses for a polygenic score (PGS) for educational attainment (EA). @molly_przew 1/n One is as a control or instrumental variable that helps better evaluate the effects of social interventions. We do not focus on this use, partly because it is farther from our area of expertise. It seems like it may be quite helpful, though of course not without assumptions. 2/n

.@molly_przew and I wrote a review of @kph3k’s book, The Genetic Lottery: Why DNA matters for social equality, for @journal_evo: [pdf: gcbias.files.wordpress.com/2022/01/geneti… ] 1/6 As its title suggests, the book makes a pitch for the centrality of genetics in understanding and addressing social inequalities. In doing so, however, we think it mischaracterizes key points as to where human genetics stands & what it promises. 2/6

Small children are quite capable of reaching their noses and yet choose not to blow them. Why? 1/n Most other primates, with some exceptions, have almost no nose. Yet human children have evolved particularly large noses. [image credit en.wikipedia.org/wiki/Proboscis… ]

I want to take a moment to talk about a Motte and Bailey tactic that often gets disingenuously used to frame and restrict conversations around human differences among groups. 1/n It's an old trick where folks try to persuade you of the following logic: if you don't buy into hereditarian arguments about behavioural differences among groups, then you must think that genetics makes zero contribution to differences among groups. 2/n

TFW I deeply regret not adding talks/classes to my UCD info page as I do them, instead of trying to work out three years later what I did when. Currently finding out when people graduated/left the group by searching Twitter for congratulations/lab party pic tweets. Like a Boss.

Bee killer (Mallophora) at park. Thought it was dead so I turned them over for a better shot, quite surprised when it flew straight past my ear. Crossing the west by train is stunning (if tiring)

https://twitter.com/Graham_Coop/status/1418646375022178315

I listened to an interview with the CEO of this new polygenic score & IVF embryo selection company (Orchid Health). It was not reassuring. 1/ mendelspod.com/podcasts/orchi… There was a lot of hype about genome sequencing and claims that healthcare workers are being paternalistic and acting as gatekeepers to information that parents deserve to have. 2/

You and I are related to each other, but the question of how closely we are related has two very different answers. One genealogical, one genetic. 1/n To see why, consider a simulation of my family tree back over the generations, my number of genealogical ancestors doubles every generation. I soon have a lot of genealogical ancestors 2/n

Updated version of @DocEdge85’s preprint on the potential for upload attacks on genetic genealogy databases, detailing various hacks. Big addition, we did a proof-of-concept demo of a simple method for obtaining peoples’ genotypes from GEDmatch 1/n
biorxiv.org/content/10.110… @DocEdge85 GEDmatch seems to use long runs of compatible genotypes between individuals (kits) to call genetic relatedness, identity by state (phase-unaware IBS calling). These runs of IBS can broken by a single opposite-homozygous genotype, e.g see user videos 2/n

I've come to the view that genetic-genealogy databases that allow uploads are a substantial risk to genetic privacy & privacy more generally. It's not just law enforcement who might be using these databases to snoop on us. There are companies that offer private citizens access to law enforcement-style tools for collecting genetic data. See article by @sarahzhang . There's no real way of knowing these may have already been used on people w.o. their permission. theatlantic.com/science/archiv…

Note that our most recent genealogical common ancestor was one of the millions of people alive at that time, and was no one special.

https://twitter.com/jordanmantha/status/1180147511715086338

I can claim that the last genealogical common ancestor of humanity lived in Africa or East Asia, and have a similar chance of being right. 2/n

A hundred years of artificial selection (red & blue indicating up & down selection respectively). Original datasets from ideals.illinois.edu/handle/2142/35… (incredible that they have individual-level data from back to 1896). My code here: github.com/cooplab/popgen…
Based on ggridges egs from @ClausWilke cran.r-project.org/web/packages/g…

So beyond speculating wildly about eugenic selection of embryos for high IQ, it's worth worrying about what Hsu's company is doing today (genomicprediction.com). They’re providing polygenic scores aimed at embryo selection for the following traits 1/n From the papers they reference (biorxiv.org/content/10.110…) they’re constructing these polygenic scores based on LASSO. A standard way to choose a sparse set of SNPs in human genetics