The pairwise mismatch distribution for molecular marker loci is a commonly used statistic in ecology to investigate historical population growth, selection and effects of demographical processes. Since the statistical properties of the stationary mismatch distribution are not yet well understood for complex models, we investigate its behavior under the neutral Wright-Fisher model with mutation, recombination, migration and local stochastic propagation of haplotypes corresponding to microepidemics in a host population. Under such circumstances the traditional analytical tools for deriving equilibrium distributions do not apply, making derivation of recursive expressions elusive. Using simulations we show that the mismatch distribution of a population typically exhibits a considerable level of stochasticity over generations unless the mutation rate is sufficiently high. The stationary mean mismatch distribution is a more stable summary but it may remain insensitive to changes in the model parameters, suggesting that additional complementary population summaries are needed for inference. We illustrate mismatch distribution dynamics using data on asymptomatic pediatric carriage of Streptococcus pneumoniae, indicating a thoroughly mixing population with rapid dynamics across local communities.
|Publication status||Submitted - 2016|
|MoE publication type||A1 Journal article-refereed|
Fields of Science
- 112 Statistics and probability
- 1183 Plant biology, microbiology, virology
Shubin, M., Numminen, S. E., Gutmann, U. M., Hanage, W. P., & Corander, J. I. (2016). Statistical properties of the allelic mismatch distribution in neutrally evolving haploid populations. Manuscript submitted for publication.