Sweptaway – Part 1

Brace yourselves for a series of new posts on selection, especially with articles from the special Molecular Ecology issue on “Detecting selection in natural populations: making sense of genome scans and towards alternative solutions” starting to roll out!

Selective sweeps (i.e reduction in genomic diversity at regions linked to positively selected, and fixed mutations in a population – also see my previous post here) are commonly identified by quantifying polymorphism at linked sites. Commonly used methods to characterize sweeps include tests with Tajima’s D, Fay and Wu’s H, the HKA chi-squared statistic, etc. Likelihood-based frameworks to infer selective sweeps include SweepFinder (Nielsen et al. 2005), SweeD (Pavlidis et al. 2011), XP-CLR (Chen et al. 2010), among others. However, reduction in genomic diversity need not necessarily be due to selective sweeps – alternate explanations could include population bottlenecks, background selection (see my earlier post explaining this), and unusually slow mutation rates. Huber et al. (2015) in their recent paper on “Detecting recent selective sweeps while controlling for mutation rate and background selection” describe an addition to Nielsen et al. (2005)’s SweepFinder.

To control for mutation rate (and possibly selective constraints), Huber et al. (2015) suggest the inclusion of invariant, fixed sites (with respect to an outgroup) in the analyses. Simulation analyses suggest increases in the power of the Composite Likelihood Ratio (CLR) test, and decrease in False Positive Rates (FPR) in detecting sweeps, with this inclusion of fixed differences. Similarly, controlling for background selection (by inclusion of a B-value map) showed greater power in detecting sweeps while including all sites i.e. polymorphic, and fixed invariant sites in the analyses.

Using the reduction in diversity relative to divergence as a necessary hallmark of a selective sweep in our model also helps to reduce false positives, e.g. in the case of a recent population bottleneck.

Nucleotide diversity across the X chromosome in CEU, YRI, and JPT human populations, compared with regions of Neandertal introgression. Image courtesy: Dutheil et al. (2015) doi/10.1371/journal.pgen.1005451.g005

In a neat application of detecting selective sweeps, Dutheil et al. (2015) also offer an alternate explanation to the observed levels of reduced diversity and divergence in human X chromosomes (with respect to Chimpanzees) to the controversial hypothesis of Patterson et al. (2006). X chromosomes have smaller effective population sizes (due to hemizygosity in males) compared to autosomes, and thus expected to be drifting more, and show lower divergence between humans and chimpanzees. By studying the patterns of incomplete lineage sorting (ILS) along X chromosome by simulating gene trees and estimating demographic parameters under a divergence model (CoalHMM), Dutheil et al. (2015) report a bimodal distribution for ILS along the X chromosome, with 8 regions identified as exhibiting particularly low proportions of ILS. Analyses of the effect of background selection on ILS showed that only about 31% of X chromosome windows could be explained solely due to background selection, whereas comparison of diversity across the X chromosome across different human populations, and the neandertal showed that low ILS regions predominantly evolve by recurrent selective sweeps. They argue that the observed large-scale reductions in diversity in extant human populations are also not plausible under a model of secondary contact between humans and chimpanzee ancestors (as suggested by Patterson et al. 2006).

Whatever the underlying mechanism, our observations demonstrate that the evolution of X chromosomes in the human chimpanzee ancestor, and in great apes in general, is driven by strong selective forces. The striking overlap between the low-ILS regions we have identified and the Neandertal introgression deserts identified by Sankararaman et al. further hints that these forces could be driving speciation.

References:
Huber, Christian D., et al. “Detecting recent selective sweeps while controlling for mutation rate and background selection.” Molecular Ecology (2015). DOI: http://dx.doi.org/10.1111/mec.13351

Dutheil, Julien Y., et al. “Strong selective sweeps on the X chromosome in the human-chimpanzee ancestor explain its low divergence.” PloS Genetics (2015). DOI: http://dx.doi.org/10.1371/journal.pgen.1005451

Chen, Hua, Nick Patterson, and David Reich. “Population differentiation as a test for selective sweeps.” Genome research 20.3 (2010): 393-402. DOI: http://dx.doi.org/10.1101/gr.100545.109

Pavlidis, Pavlos, et al. “SweeD: likelihood-based detection of selective sweeps in thousands of genomes.” Molecular biology and evolution (2013): mst112. DOI: http://dx.doi.org/10.1093/molbev/mst112

Nielsen, Rasmus, et al. “Genomic scans for selective sweeps using SNP data.”Genome research 15.11 (2005): 1566-1575. DOI: http://dx.doi.org/10.1101/gr.4252305

Sankararaman, Sriram, et al. “The genomic landscape of Neanderthal ancestry in present-day humans.” Nature 507.7492 (2014): 354-357. DOI: http://dx.doi.org/10.1038/nature12961

Patterson, Nick, et al. “Genetic evidence for complex speciation of humans and chimpanzees.” Nature 441.7097 (2006): 1103-1108. DOI: http://dx.doi.org/10.1038/nature04789

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 adaptation, evolution, genomics, mutation, population genetics, selection, speciation, theory and tagged , , , , , . Bookmark the permalink.