Divergence and Linked Background Selection

We have widely discussed the reduction in neutral diversity due to demography and linked selection effects (e.g. selective sweeps and hitchhiking, or background selection) in several previous posts (e.g see here, here, and here). However, how linked selection affects neutral divergence between species isn’t as clear, specifically due to difference in divergence times between species. For e.g. reduction in neutral diversity in the ancestor (e.g. due to background selection) theoretically would be expected to have negligible effect on current species (due to new mutations since the split) in long-diverged species. How then are patterns of divergence affected due to ancestral linked selection at linked neutral sites?  Phung et al. (2016) address this question using simulations and examining neutral divergence between species with a wide range of divergence times – humans versus chimpanzees, orangutans, and mice.

phung2016

Models of the effects of linked background selection on species with different divergence times (short versus long). Red indicates coalescent times at neutral sites, and blue lines show divergence times at linked neutral sites affected by background selection. Image courtesy Phung et al. 2016 http://journals.plos.org/plosgenetics/article?id=info:doi/10.1371/journal.pgen.1006199

Testing the hypothesis that if natural selection indeed reduces divergence at linked neutral sites, then this effect would be greater at functional regions, Phung et al. 2016 find a negative correlation between divergence and functional content, and a positive correlation between human-primate neutral divergence and recombination rate. This indicates that neutral divergence is reduced at regions with greater functional constraint, and more tightly linked to those under the direct effect of selection (low recombination ~ low divergence, low divergence ~ greater functional constraint). A similar test between human and mouse genomes reveals the same negative correlation between divergence and functional content, and a positive correlation between divergence and strength of background selection. Phung et al. (2016) also recapitulate these patterns of correlation between divergence and recombination and background selection using simulations. Additionally, they also propose a simple two-locus model to explain these correlations between species with long divergence times (sensu humans and mice).

The authors however also point to alternate scenarios which can also contribute to the same/similar patterns of correlations:

  • Selective sweeps, and hitchhiking
  • Direct purifying selection effects
  • Variation in mutation rates across the genome
  • Biased gene conversion

The authors also echo the now-oft-cited sentiment about accounting for the effects of linked selection while using methods for demographic inference (also see my post on Schrider et al. 2016 here).

Our finding that background selection can increase the variance in coalescent times across the genome suggests these methods as well as other statistical methods which seek to infer demographic history from the distribution of coalescent times across the genome, such as the PSMC approach, should account for the increased variance in coalescent times across the genome due to background selection. Not accounting for background selection could result in inferring spurious demographic events to account for the additional variance in coalescent times across the genome as has recently been suggested for positive selection

Reference:

Phung, Tanya N., Christian D. Huber, and Kirk E. Lohmueller. “Determining the Effect of Natural Selection on Linked Neutral Divergence across Species.” PLoS Genet 12.8 (2016): e1006199. DOI: http://dx.doi.org/10.1371/journal.pgen.1006199

Schrider, Daniel, Alexander G. Shanku, and Andrew D. Kern. “Effects of linked selective sweeps on demographic inference and model selection.”bioRxiv (2016): 047019. DOI: http://dx.doi.org/10.1101/047019

Share

About Arun Sethuraman

I am a computational biologist, and I build statistical models and tools for population genetics. I am particularly interested in studying the dynamics of structured populations, genetic admixture, and ancestral demography.
This entry was posted in evolution, genomics, methods, selection, speciation, theory and tagged , , , , , . Bookmark the permalink.