I'm currently a postdoc with Wolfgang Stephan, funded by a DFG grant through the Schwerpunkt programme SPP 1819 "Rapid Evolutionary Adaptation". We are located at the Museum für Naturkunde in Berlin. Previously I was a postdoc with Jochen Blath at TU Berlin; funded by a DFG grant (to Jochen Blath and Matthias Birkner at JGU Mainz) through the Schwerpunkt programme SPP 1590 "Probabilistic Structures in Evolution". From September 2009 until October 2012, I was postdoc at the Department of Statistics at University of Oxford working with Alison Etheridge. John Wakeley at Harvard University supervised my PhD work on gene genealogies and large offspring numbers.
Museum Office: N 1210
Museum phone number: 0049 302 093 70379
Museum Email: bjarki.eldon 'at' mfn-berlin.de
TU Berlin very generously allows me to keep my office and computer account at the Math Institute:
TU Berlin phone number: 0049 30 314 25762
TU Berlin Email: eldon 'at' math.tu-berlin.de
Research interests: Population genetics, evolutionary biology, population models which admit high fecundity and skewed offspring distributions, multiple merger coalescent processes, marine genomics, probability theory
Grant of Excellence: Population genomics of highly fecund codfish Icelandic Research Fund Grant of Excellence awarded jointly with Einar Arnason (corresponding PI), and with PIs Katrin Halldorsdottir, Alison Etheridge, and Wolfgang Stephan.
DFG SPP 1819 Start-up module grant: Grant awarded jointly with Jere Koskela and Maite Wilke-Berenguer:Population genomics of highly fecund codfish.
On my research: Some marine organisms like Atlantic cod (Gadus morhua) and Pacific oysters (Crassostrea gigas) have high fecundity to make up for high early mortality. They also tend to exhibit a large number of genetic variants present in low-copy numbers. These characteristics imply that, occasionally at least, a few lucky parents have very many offspring. Population models traditionally employed are applicable to populations with low numbers of offspring; humans are a good example of such a population. I'm interested in population models that allow for large numbers of offspring. The coalescent processes that result from these models admit multiple and simultaneous multiple mergers of ancestral lineages, and so are special cases of the Lambda- and Xi-coalescents introduced by Donnelly and Kurtz (1999), Pitman (1999), Sagitov (1999), and Schweinsberg (2000). A lot of work remains in understanding large offspring numbers in terms of predictions about genetic diversity and developing inference methods. Apart from a pure academic interest, better understanding of the genetics of commercially important marine populations should improve conservation and management efforts. Multiple merger coalescent models may also find applications in medical genetics and epidemiology.
Natural populations interact in many ways. Two examples are predation, and competition for resources such as space. Developing population models, and the resulting coalescent processes, that account for interactions among populations is a major goal. Developing and classifying multiple merger coalescent processes that allow for stochastically changing population size would be a step in that direction.Matthias Birkner, Jochen Blath and I have derived a multi-loci ancestral recombination graph admitting simultaneous mergers, which will be useful for inferring between multiple merger coalescent models using multi-loci data. Expected values of the site-frequency spectrum associated with Lambda-coalescents are also very helpful when inference is based on single loci data.
With Fabian Freund we have shown that, at least some multiple-merger coalescent models, can be distinguished from exponential population growth.
James Degnan, Joe Zhu and I are studying multiple merger coalescent models within complex species trees. Joe (with a bit of input from James and me) is developing Hybrid-Lambda, which can simulate gene genealogies with multiple mergers within species trees, generate data and compute various statistics. These calculations will be important for phylogeny reconstruction of high fecundity populations with skewed offspring distributions.