Substack cometh, and lo it is good. (Pricing)

Between the tree & ring

The_history_and_geography_of_human_genes_Luigi_Luca_Cavalli-Sforza_map_genetic
Credit: History and Geography of Human Genes

PhylogeneticTree“Tree thinking”, just like “population thinking”, is essential to understanding evolutionary biology. But there are problems with this. First, even on a macroevolutionary scale there is massive violation of separation between the branches of the tree of life due to lateral gene flow, whether directly or mediated via viruses. As you drill down to a finer phylogenetic grain the issue of reticulation, the transformation of a tree into a graph, becomes something you have to integrate into your model. This is one reason that TreeMix was developed, it allowed one to model gene flow across the branches of trees on a microevolutionary scale. But there’s another major issue, and that is of clines. The “graph” generated by TreeMix still models pulse admixture events. But a lot of genetic variation is generated in the context of isolation by distance dynamics. There are no singular events where populations mix, rather, every instance is part of a continuous process of mixing.* An excellent illustration of this scenario is a ring species.

When generating a “tree of life” it seems that it behooves us to take these details into account, because to a great extent the details are really what we’re interested in now (OK, at least if we’re population geneticists). Each species or section of the tree of life may exhibit different local dynamics, just like the topography of our planet exhibits local variation. Some regions may be subject to far greater gene flow across the branches rather far from the tips (e.g., plants), while others exhibit rather less (e.g., some mammalian lineages). As far as humans go, about 10 years ago a hybrid tree-cline model seemed viable, as outlined in Ramachandran et al., a serial founder event out of Africa, and then equilibration with isolation by distance, leading to a clinal overlay upon the branching process. After Pickrell et al. I think that this model just can’t suffice. Rather, there were powerful pulse admixture events due to meta-population dynamics and large scale demographic implications of cultural change over the last 10,000 years, which wreaked havoc with patterns of human genetic variation. As on the larger tree of life the dynamics which characterized particular segments of the human phylogenetic tree/graph vary. In the New World Ramachandran et al.’s original formulation might actually be rather good south of the Rio Grande. In contrast it just doesn’t work very well in South and Southeast Asia, which has been subject to a great deal of genetic change over the past ~10,000 years, well after the Out of Africa event (in these two cases there look to be pulse fusions between very distinct branches of the human phylogenetic tree in recent history).

These thoughts were stimulated in part by comments over at 3 Quarks Daily [link fixed] in response to an Omar Ali post where he mirrored content from this weblog. This elicited two broad reactions. First, some readers objected to Omar posting content from me because I’m a conservative who holds views taboo to the liberal mainstream of that website. Because obviously that’s what being liberal is, knowing what to believe, and not giving any voice to beliefs and viewpoints outside of what you label to be right, true, and orthodox.** But a more interesting objection is to those who think that my adherence to the race concept is not supported by science. Here I feel like I’m talking to Creationists, they know what they believe, but they don’t know much about what they don’t believe. Most Creationists don’t really know much about evolutionary biology, they just know their talking points. Similarly, many who object to my acceptance of the validity of the race concept just trot out talking points. Intriguingly there is a similarity to the objections by anti-evolution Creationists and anti-population structure Creationists: a fixation on Platonic categories/kinds. A major confusion that anti-evolution Creationists exhibit is that species, “kinds”, are clear and distinct categories. They’re not. So many of the critiques fail at the get-go, because when you reject a Platonic idea of species the rejection of anti-evolution Creationism is almost axiomatic. The same aspect of disagreement emerges when arguing with anti-population structure Creationists. First, they don’t know the literature, so their objections are often weird and ad hoc. That’s fine. But the bigger problem is that I don’t hold the Platonic model of population structure they seem to think I must hold. Rather, our terms or categories are only useful in an instrumental sense. Their validity or lack thereof reflects how well they model the real processes which shape the genealogies which collectively define particular populations. I can try to get this viewpoint across, but since I hold views which disagree with their views (or so they think) and so naturally am wrong, I generally don’t make much headway.

* This is where Gideon Bradburd’s SpaceMix should help. The preprint will be out soon….

** This is a joke. I’m aware there are many liberal readers of this website. But, you must admit that it’s pretty funny how narrow-minded people who label themselves “liberal” can be! E.g., the fact that I’m a self-described conservative is reason enough not to give voice to my ideas.

Posted in Uncategorized

Comments are closed.