I'm a postdoc in the Fred Adler lab, and my research focuses on several aspects of evolutionary theory. Currently, I'm most interested in models of coalescence that include variability in fitness among individuals. We currently lack a theory that can accurately predict how population size and selection combine to shape the structure of gene genealogies and the resulting patterns of genetic diversity. Simulations indicate that even very small levels of heritable variation in reproductive success can significantly influence patterns of genetic variation in a population. I'm presently working on a model that imagines fitness as a quantitative genetic trait describing expected reproductive success. Given some description of the mutational process, as well as a guess regarding the population distribution of fitnesses, one can make some predictions regarding the distribution of pairwise coalescent times and and the expected lengths of coalescent intervals in a genealogy. Most interestingly, despite the fact that the function describing the change in expected reproductive success is very complex in reality, simulations suggest that the results are influenced only by the variance of the function, and not any other property. This finding suggests that very simple analytic models may accurately capture the effects of much more realistic and complex descriptions of mutation, assuming the variance in fitness differences is roughly equal.
In addition to my recent interest in the coalescent, I also spend a fair amount of time thinking about the interaction between natural selection, drift, and basic properties of genomes, such as the distribution of mutational effects. Why is the average mutation so greatly deleterious for some species (such as RNA viruses), but relatively mild for us metazoans? How do factors like population structure and the mutation rate affect the mean and variance of the distribution? Last year, Fred, Steven Proulx, and I published one paper demonstrating that in some cases natural selection may even favor more negative mean for the distribution, suggesting that in some cases it may be 'better' for deleterious mutations to have as severe an impact as possible! (O'Fallon et al. 2007. Quasispecies evolution in structured populations favors maximally deleterious mutations. Proc. R. Soc. B 274)
When not pondering the mysteries of evolution, or staring at long lists of computer generated numbers (these appear to be roughly equivalent on occasion), I'm probably climbing (summer, fall) or skiing (winter, spring).
O'Fallon, B. Two optimal mutation rates and the evolution of virulence in obligate pathogens. Manuscript in prep
O'Fallon, B., F.R. Adler, S. R. Proulx. 2007. Quasi-species evolution in subdivided populations favors maximally deleterious mutations. Proc. Roy. Soc. London B. 274:3159-3164
O'Fallon, B. 2007. Population structure, levels of selection, and the evolution of intracellular symbionts. Evolution 62(2):361-373 O'Fallon, B. & F. Adler. 2006.
Stochasticity, complex spatial structure, and the feasibility of the shifting balance theory. Evolution 60(3):448-59
Brown, J. M., M. Todd-Thompson, A. McCord, A. O'Brien, B. O'Fallon. 2006. Phylogeny, host association, and wing pattern variation in the endemic hawaiian fruit ies (Diptera, Tephritidae). Instrumenta Biodiversitatus VII:1-16
Photo: Atop Mt. Wilson, Red Rocks, Nevada, after completing The Resolution Arete (V 5.10 A0, Broscocak & Conley 1981), with Matt Israel
Feel free to contact me at