pubmed.ncbi.nlm.nih.gov

Inferring the joint demographic history of multiple populations from multidimensional SNP frequency data - PubMed

Inferring the joint demographic history of multiple populations from multidimensional SNP frequency data

Ryan N Gutenkunst et al. PLoS Genet. 2009 Oct.

Abstract

Demographic models built from genetic data play important roles in illuminating prehistorical events and serving as null models in genome scans for selection. We introduce an inference method based on the joint frequency spectrum of genetic variants within and between populations. For candidate models we numerically compute the expected spectrum using a diffusion approximation to the one-locus, two-allele Wright-Fisher process, involving up to three simultaneous populations. Our approach is a composite likelihood scheme, since linkage between neutral loci alters the variance but not the expectation of the frequency spectrum. We thus use bootstraps incorporating linkage to estimate uncertainties for parameters and significance values for hypothesis tests. Our method can also incorporate selection on single sites, predicting the joint distribution of selected alleles among populations experiencing a bevy of evolutionary forces, including expansions, contractions, migrations, and admixture. We model human expansion out of Africa and the settlement of the New World, using 5 Mb of noncoding DNA resequenced in 68 individuals from 4 populations (YRI, CHB, CEU, and MXL) by the Environmental Genome Project. We infer divergence between West African and Eurasian populations 140 thousand years ago (95% confidence interval: 40-270 kya). This is earlier than other genetic studies, in part because we incorporate migration. We estimate the European (CEU) and East Asian (CHB) divergence time to be 23 kya (95% c.i.: 17-43 kya), long after archeological evidence places modern humans in Europe. Finally, we estimate divergence between East Asians (CHB) and Mexican-Americans (MXL) of 22 kya (95% c.i.: 16.3-26.9 kya), and our analysis yields no evidence for subsequent migration. Furthermore, combining our demographic model with a previously estimated distribution of selective effects among newly arising amino acid mutations accurately predicts the frequency spectrum of nonsynonymous variants across three continental populations (YRI, CHB, CEU).

PubMed Disclaimer

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Figure 1
Figure 1. Frequency spectrum gallery.

(A) Qualitative effects of modeled neutral genetic forces on formula image, the density of alleles at relative frequencies formula image and formula image in populations 1 and 2. (B) For the spectra shown, an equilibrium population of effective size formula image diverges into two populations formula image generations ago. Populations 1 and 2 have effective sizes formula image and formula image, respectively. Migration is symmetric at formula image per generation, and formula image. (C) The AFS at formula image. Each entry is colored by the logarithm of the number of sites in it, according to the scale shown. (D) The AFS at various times for various demographic parameters, on the same scale as (B). (E) Comparison between coalescent- and diffusion-based estimates of the likelihood formula image of data generated under the model (A). Coalescent-based estimates of the likelihood, each of which took approximately 7.0 seconds, are represented in the histogram. The result from our diffusion approach, which took 2.0 seconds, is represented by the red line. For accuracy comparison, the yellow line indicates the likelihood inferred from formula image coalescent simulations.

Figure 2
Figure 2. Out of Africa analysis.

(A) AFS for the YRI, CEU, and CHB populations. The color scale is as in (C). (B) Illustration of the model we fit, with the 14 free parameters labeled. (C) Marginal spectra for each pair of populations. The top row is the data, and the second is the maximum-likelihood model. The third row shows the Anscombe residuals between model and data. Red or blue residuals indicate that the model predicts too many or too few alleles in a given cell, respectively. (D) The observed decay of linkage disequilibrium (black lines) is qualitatively well-matched by our simulated data sets (colored lines). (E) Goodness-of-fit tests based on the likelihood formula image and Pearson's formula image statistic both indicate that our model is a reasonable, though incomplete description of the data. In both plots, the red line results from fitting the real data and the histogram from fits to simulated data. Poorer fits lie to the right (lower formula image and higher formula image). (F) The improvement in likelihood from including contemporary migration in the real data fit (red line) is much greater than expected from fits to simulated data generated without contemporary migration (histogram). This indicates that the data contain a strong signal of contemporary migration.

Figure 3
Figure 3. Settlement of the New World analysis.

As in Figure 2, (A) is the data, (B) is a schematic of the model we fit, (C) compares the data and model AFS, and (D) compares LD. (E) The fit of our model to the real data is not atypical of fits to simulated data. (F) The improvement in real data fit upon including CHB-MXL migration (red line) is very typical of the improvement in fits to simulated data without CHB-MXL migration. Thus we have no evidence for CHB-MXL migration after divergence.

Figure 4
Figure 4. Distribution of nonsynonymous polymorphism.

We simulated our maximum-likelihood Out of Africa demographic model with a distribution of selective effects previously inferred for nonsynonymous polymorphism . (A) To enable direct comparison with the neutral AFS (Figure 2C), the scaled mutation rate formula image was set identically, as is the color scale. As expected, selection dramatically reduces the amount of segregating polymorphism. (B) Shown are the proportions of variation found in various frequency classes. As expected, nonsynonymous variants typically have lower frequency. They also less likely to be shared between populations. Data error bars indicate 95% bootstrap confidence intervals.

Similar articles

Cited by

References

    1. Mellars P. Going east: new genetic and archaeological perspectives on the modern human colonization of Eurasia. Science. 2006;313:796–800. - PubMed
    1. Goebel T, Waters MR, O'Rourke DH. The late Pleistocene dispersal of modern humans in the Americas. Science. 2008;319:1497–1502. - PubMed
    1. Nielsen R, Hellmann I, Hubisz M, Bustamante C, Clark AG. Recent and ongoing selection in the human genome. Nat Rev Genet. 2007;8:857–868. - PMC - PubMed
    1. Adams AM, Hudson RR. Maximum-likelihood estimation of demographic parameters using the frequency spectrum of unlinked single-nucleotide polymorphisms. Genetics. 2004;168:1699–1712. - PMC - PubMed
    1. Marth GT, Czabarka E, Murvai J, Sherry ST. The allele frequency spectrum in genome-wide human variation data reveals signals of differential demographic history in three large world populations. Genetics. 2004;166:351–372. - PMC - PubMed

Publication types

MeSH terms