My Research
I am interested in population structure - by which I mean anything that causes a population to deviate from the nice, simple and mostly unrealistic model of complete random mating between all members of the population. Under this very general definition it becomes clear that almost everything interesting in biology leads to population structure of some kind; including habitat fragmentation, competition for resources, mate choice etc.
Perhaps the most obvious way in which structure can arise is by isolation by distance, meaning that individuals tend to mate with those members of the population that are geographically close to them. This makes intuitive sense as all animals are limited in the range that they can move within a lifetime, and are forced to make all other choices (such as choice of "the best" mate) within this overarching constraint of availability. The net effect is that individuals tend to be more genetically similar to other individuals that are close by compared with those that are far away.

The mathematical analysis of population structure is made easier by a number of powerful and relatively recent discoveries within population genetics. First and foremost is the work of pioneers such as Wright, Malecot and Fisher, who laid the groundwork by coming up with a mathematical formulation of the process of genetic drift (the Wright-Fisher model). This is complemented by Kingman's n-coalescent, known more generally as coalescent theory, which allows us to take a retrospective view of the process of drift that may have produced the sample that we see today. There have been a great many discoveries in coalescent theory since its inception, with the invariance principle of Charlesworth et al. being of particular importance as it enables us to get to grips with models of migration between subpopulations.

However, even with these tools at our disposal it seems a daunting task to model anything but the simplest scenarios. Perhaps as a consequence of this there has been a tendency within population genetics to try to boil down all the complexity of a situation in to a single all-powerful statistic (traditionally Wright's Fst). While this trust in statistics may not be entirely misplaced (in many biologically meaningful scenarios a single statistic can tell us a great deal), it can also be a dangerous distraction - there have been a great many studies that have come to conclusions based on a flawed understanding of Fst. As such, another of the main aims of my research is to get to grips with the strengths and weaknesses of different statistical philosophies, including Frequentist, Bayesian, pure Likelihood, and just counting! This becomes particularly important in light of second generation sequencing, which is sure to expose any Elephants in the room.

Well, I've said rather more than I intended, but hopefully some of it will have been interesting to some of you! If anyone wants to talk more about any of these things, or in fact about anything, then feel free to check out my website, or get in touch.

Looking forward to hearing what other people are up to.