Figure 3. Inference of admixture in sub-Saharan Africa using MALDER.
We used MALDER to identify the evidence for multiple waves of admixture in each population. (A) For each population, we show the ancestry region identity of the two populations involved in generating the MALDER curves with the greatest amplitudes (coloured blocks) for at most two events. The major contributing sources are highlighted with a black box. Populations are ordered by ancestry of the admixture sources and dates estimates which are shown 1.96 s.e. For each event we compared the MALDER curves with the greatest amplitude to other curves involving populations from different ancestry regions. In the central panel, for each source, we highlight the ancestry regions providing curves that are not significantly different from the best curves. In the Jola, for example, this analysis shows that, although the curve with the greatest amplitude is given by Khoesan (green) and Eurasian (dark yellow) populations, curves containing populations from any other African group (apart from Afroasiatic) in place of a Khoesan population are not significantly smaller than this best curve (SOURCE 1). Conversely, when comparing curves where a Eurasian population is substituted with a population from another group, all curve amplitudes are significantly smaller (). (B) Comparison of dates of admixture 1.96 s.e. for MALDER dates inferred using the HAPMAP recombination map and a recombination map inferred from European (CEU) individuals from Hinch et al. (2011). We only show comparisons for dates where the same number of events were inferred using both methods. Point symbols refer to populations and are as in Figure 1. (C) as (B) but comparison uses an African (YRI) map. Source data can be found in Figure 3—source data 1.
DOI: http://dx.doi.org/10.7554/eLife.15266.014
Figure 3—source data 1. The evidence for multiple waves of admixture in African populations using MALDER and the HAPMAP recombination map.
For each event in each ethnic group we show the largest inferred amplitude and date of an admixture event involving two reference populations (Pop1 and Pop2). We additionally provide the ancestry region identity of the two main reference populations, together with scores for curve comparisons between this best curve and those containing populations from different ancestry regions. We use a cut-off of 2 to decide whether sources from multiple ancestries best describe the admixture source.
elife-15266-fig3-data1.xlsx (59.5KB, xlsx)
DOI: 10.7554/eLife.15266.015
Figure 3—source data 2. The evidence for multiple waves of admixture in African populations using MALDER and the African recombination map.
Columns as in Figure 3—source data 1.
elife-15266-fig3-data2.xlsx (59.8KB, xlsx)
DOI: 10.7554/eLife.15266.016
Figure 3—source data 3. The evidence for multiple waves of admixture in African populations using MALDER and the European recombination map.
Columns as in Figure 3—source data 1.
elife-15266-fig3-data3.xlsx (60KB, xlsx)
DOI: 10.7554/eLife.15266.017
Figure 3—source data 4. The evidence for multiple waves of admixture in African populations using MALDER and the HAPMAP recombination map and a mindis of 0.5cM.
Columns are as in Figure 3—source data 1. Here we show the results for the MALDER analysis where we over-ride any short-range LD and define a minimum distance of 0.5cM from which to start computing admixture LD curves
elife-15266-fig3-data4.xlsx (60.6KB, xlsx)
DOI: 10.7554/eLife.15266.018
Figure 3—figure supplement 1. Weighted LD amplitudes for a selection of 9 ethnic groups.
For a given test population we show the amplitude ( 1 s.e.) computed using a test population and every other population as the second reference. Plotted are the fitted amplitudes for each set of curves with the population used labelled beneath, with populations ordered by amplitude. A large number of population showed a similar profile to (A), that is, with Eurasian populations showing the highest amplitudes. Other populations, e.g. Malawi, obtained the largest amplitudes from an African population.
Figure 3—figure supplement 2. Comparison of weighted LD amplitude scores across all African ethnic groups.
For a given test population we computed the ALDER amplitude (y-axis intercept) using the test population and every other population as the second reference. We then ranked the amplitudes across a given test population: populations who gave the top-ranked (i.e. largest) amplitude are in green, with those beneath a rank 15 shown in grey. This analysis shows that for many populations the reference populations giving the largest amplitudes (i.e. have the highest rank) are often non-African groups.
Figure 3—figure supplement 3. Comparison of the minimum distance to begin computing admixture LD.
For each of the 48 African populations as a target, we used ALDER to compute the minimum distance over which short-range LD is shared with each of the 47 other African and 12 Eurasian reference populations. Here we show boxplots showing the distribution of minimum inferred genetic distances (y-axis) over which LD is shared for each of the reference populations separately (x-axis). We performed two analyses using weighted LD, one using these values of the minimum distance inferred from the data, and another where this distance was forced to be 0.5cM (dotted red line). Across all African populations we observe LD correlations with other African populations at genetic distances 0.5cM, with median values ranging between 0.7cM when GUMUZ is used as a reference to 1.4cM when FULAII is used as a reference. In fact, when we further explore the range of these values across each region separately (Figure 3—figure supplement 3), we note that, as expected, these distances are greater between more closely related groups.
Figure 3—figure supplement 4. Comparison of the minimum distance to begin computing admixture LD split by region.
As in Figure 3—figure supplement 3 except distances are stratified by region. The median minimum distance that all sub-Saharan African populations have correlated LD is always greater than 0.5cM. Taken together with the results described in Figure 3—figure supplement 4, this suggests that all African populations share some LD over short genetic distances, that may be the result of shared demography or admixture. (Note that ALDER computes LD correlations at distances 2cM.)
Figure 3—figure supplement 5. Results of MALDER for all populations using a European specific recombination map.
We used MALDER to identify the evidence for multiple waves of admixture in each population. (A) For each population, we show the ancestry region identity of the two populations involved in generating the MALDER curves with the greatest amplitudes (which are the closest to the true admixing sources amongst the reference populations) for at most two events. The sources generating the greatest amplitude are highlighted with a black box. Populations are ordered by ancestry of the admixture sources and dates estimates which are shown 1 s.e. (B) Comparison of dates of admixture 1 s.e. for MALDER dates inferred using the HAPMAP recombination map and a recombination map inferred from European (CEU) individuals from Hinch et al. (2011). We only show comparisons for dates where the same number of events were inferred using both methods. Point symbols refer to populations and are as in Figure 1. (C) as (B) but comparing with an African (YRI) map.
Figure 3—figure supplement 6. Results of the MALDER analysis computing weighted admixture decay curves from 0.5cM.
As in the main analyses, the algorithm was run independently three times with the HAPMAP, YRI, and CEU genetic maps. The main results shown here are from the HAPMAP analysis. For each population, we show the ancestry region identity of the two populations involved in generating the MALDER curves with the greatest amplitudes (which are the closest to the true admixing sources amongst the reference populations) for at most two events. The sources generating the greatest amplitude are highlighted with a black box. Populations are ordered by ancestry of the admixture sources and dates estimates which are shown 1 s.e. (B) Comparison of dates of admixture 1 s.e. for MALDER dates inferred using the HAPMAP recombination map and a recombination map inferred from European (CEU) individuals from (Hinch et al., 2011). We only show comparisons for dates where the same number of events were inferred using both methods. Point symbols refer to populations and are as in Figure 1. (C) as (B) but comparing with an African (YRI) map.
Figure 3—figure supplement 7. Results of MALDER for all populations using an African specific recombination map.
We used MALDER to identify the evidence for multiple waves of admixture in each population. (A) For each population, we show the ancestry region identity of the two populations involved in generating the MALDER curves with the greatest amplitudes (which are the closest to the true admixing sources amongst the reference populations) for at most two events. The sources generating the greatest amplitude are highlighted with a black box. Populations are ordered by ancestry of the admixture sources and dates estimates which are shown 1 s.e. (B) Comparison of dates of admixture 1 s.e. for MALDER dates inferred using the HAPMAP recombination map and a recombination map inferred from European (CEU) individuals from Hinch et al. (2011). We only show comparisons for dates where the same number of events were inferred using both methods. Point symbols refer to populations and are as in Figure 1. (C) as (B) but comparing with an African (YRI) map.