I use mechanistic mathematical models, genomic data, and computational statistical tools to answer biological questions.


One of the subjects I am most interested in right now is the recombination rate variation1.

  • A particularly salient way recombination varies is along genomes. The landscape (i.e., probability distribution) of recombination varies between gametes, individuals, sexes, populations, and species2 and appears to be quite evolutionarily liable. A wealth of data (e.g., from single sperm sequencing in humans and other species) makes it an exciting time to think about variation in recombination at these scales.

  • One reason it is difficult to understand the causes and consequences of variation in recombination on the genome is that the phenotype is effectively infinite-dimensional. The recombination landscape on the chromosome is best thought of as a function of genomic location, with co-variances at all pairs of locations that constrain how it can evolve. With Puneeth Deraje, I’ve been working on a mechanistic model that describes how this function evolves under neutrality (i.e., drift and cis modifier accumulation). We have shown that, in a finite population, the mean CDF of recombination follows a certain Gaussian process. We are currently using comparative data to make inferences (under this model) about the tempo and mode of landscape evolution, i.e., the manner in which modifiers accumulate.


My other main scientific interest is virus evolution and emergence. RNA viruses are so rapidly evolving that, by conventional metrics of the rate of molecular evolution, their genomes should turnover neutrally in $\sim 10^3$ years. By some estimates, the rate of recombination (during co-infection of the same host) is even greater. Isn’t that remarkable! I am interested in understanding how viral mutation and recombination shape the propensity of the pathogen to successfully spillover and sustain transmission (i.e., emerge) in humans3. I am also very interested in how, as phenotypes, viral mutation and recombination rates vary and evolve.

  • Although we know recombination and reassortment are common in RNA viruses and of concern for human health, the evolution of these process is not particularly well-understood. An especially interesting alternative to the idea reassortment evolved to shuffle genetic material is that it is a necessary byproduct of the segmented nature of some viral genomes, which evolved because it allows finer control over gene expression. Weighing the shuffling and gene expression control hypotheses is a problem I am very interested in.


  1. A great introduction to the subject is this paper by Peñalba and Wolf. ↩

  2. In fact, there is a remarkable experiment showing that the gene, PRDM9, that determines where recombination occurs on the genome (i.e., the set of recombination hotspots) in many vertebrates underlies the formation of different species to begin with 😱. ↩

  3. And there happens to be a paper to this end on bioRxiv! ↩