Limits...
Unravelling the hidden ancestry of American admixed populations.

Montinaro F, Busby GB, Pascali VL, Myers S, Hellenthal G, Capelli C - Nat Commun (2015)

Bottom Line: The movement of people into the Americas has brought different populations into contact, and contemporary American genomes are the product of a range of complex admixture events.Here we apply a haplotype-based ancestry identification approach to a large set of genome-wide SNP data from a variety of American, European and African populations to determine the contributions of different ancestral populations to the Americas.Our results provide a fine-scale characterization of the source populations, identify a series of novel, previously unreported contributions from Africa and Europe and highlight geohistorical structure in the ancestry of American admixed populations.

View Article: PubMed Central - PubMed

Affiliation: 1] Institute of Legal Medicine, Catholic University, Largo F. Vito 1, Rome 00168, Italy [2] Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UK.

ABSTRACT
The movement of people into the Americas has brought different populations into contact, and contemporary American genomes are the product of a range of complex admixture events. Here we apply a haplotype-based ancestry identification approach to a large set of genome-wide SNP data from a variety of American, European and African populations to determine the contributions of different ancestral populations to the Americas. Our results provide a fine-scale characterization of the source populations, identify a series of novel, previously unreported contributions from Africa and Europe and highlight geohistorical structure in the ancestry of American admixed populations.

No MeSH data available.


Contribution of the most by informative 23 clusters inferred by fineSTRUCTURE to the analysed recipient populations.Contribution of the most informative 23 clusters to the American and Caribbean populations estimated using the non-negative least square approach. Standard error based on jack-knife resampling (22 replicates) is reported.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4374169&req=5

f3: Contribution of the most by informative 23 clusters inferred by fineSTRUCTURE to the analysed recipient populations.Contribution of the most informative 23 clusters to the American and Caribbean populations estimated using the non-negative least square approach. Standard error based on jack-knife resampling (22 replicates) is reported.

Mentions: We fit each of the American admixed populations as a mixture of the identified donor groups19 (see Methods, Supplementary Data 1). The contribution to the American admixed populations for the 23 most representative clusters and macro-areas is reported in Fig. 3 and Supplementary Fig. 1. This analysis assumes that haplotypes from the admixing populations are well represented within a mixture of present-day sampled groups. We were concerned that the demographic and evolutionary complexity of the peopling of the Americas20, coupled with the high genetic drift among Native American populations, might make the identification of the Native American contribution challenging. In particular, the true admixing groups from this region might be highly drifted from the possible ‘donor’ groups sampled, particularly given our geographically relatively sparse sample of such donor groups. To reduce this effect we always allowed a single well-sampled East Asian group (China) as a potential donor in the analysis, to act as a surrogate for haplotypes carried by any Native American donor population incompletely captured as a mixture of sampled Native American groups. Because this donor group is still likely to be strongly drifted relative to this East Asian ‘surrogate’, we also repeated our analysis after ‘masking’ direct copying of China population in the mixture-fitting step, although we still allowed all groups to contribute in the mixture. We compared the continental ancestry contributions from the full painting and the East Asian masked painting with an ADMIXTURE21 analysis performed at K=3 (Supplementary Figs 2 and 3), which closely matches the Africa, Europe and Asia/Native Americans partition. Continental ancestry estimates are highly correlated (P value <10−12) between all three approaches (Supplementary Fig. 2), although the squared distance between the masked continental ancestry estimates and that estimated by ADMIXTURE21 was, respectively, 5.4-fold and 7.9-fold reduced by the masking procedure for Europe and Asia/Native Americans, suggesting a slight gain in accuracy using this procedure. No major difference is seen for African contributions, while identified donor populations contributing to the mixture were very similar in both approaches; therefore, we henceforth report results on the basis of the masking procedure (Fig. 3).


Unravelling the hidden ancestry of American admixed populations.

Montinaro F, Busby GB, Pascali VL, Myers S, Hellenthal G, Capelli C - Nat Commun (2015)

Contribution of the most by informative 23 clusters inferred by fineSTRUCTURE to the analysed recipient populations.Contribution of the most informative 23 clusters to the American and Caribbean populations estimated using the non-negative least square approach. Standard error based on jack-knife resampling (22 replicates) is reported.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4374169&req=5

f3: Contribution of the most by informative 23 clusters inferred by fineSTRUCTURE to the analysed recipient populations.Contribution of the most informative 23 clusters to the American and Caribbean populations estimated using the non-negative least square approach. Standard error based on jack-knife resampling (22 replicates) is reported.
Mentions: We fit each of the American admixed populations as a mixture of the identified donor groups19 (see Methods, Supplementary Data 1). The contribution to the American admixed populations for the 23 most representative clusters and macro-areas is reported in Fig. 3 and Supplementary Fig. 1. This analysis assumes that haplotypes from the admixing populations are well represented within a mixture of present-day sampled groups. We were concerned that the demographic and evolutionary complexity of the peopling of the Americas20, coupled with the high genetic drift among Native American populations, might make the identification of the Native American contribution challenging. In particular, the true admixing groups from this region might be highly drifted from the possible ‘donor’ groups sampled, particularly given our geographically relatively sparse sample of such donor groups. To reduce this effect we always allowed a single well-sampled East Asian group (China) as a potential donor in the analysis, to act as a surrogate for haplotypes carried by any Native American donor population incompletely captured as a mixture of sampled Native American groups. Because this donor group is still likely to be strongly drifted relative to this East Asian ‘surrogate’, we also repeated our analysis after ‘masking’ direct copying of China population in the mixture-fitting step, although we still allowed all groups to contribute in the mixture. We compared the continental ancestry contributions from the full painting and the East Asian masked painting with an ADMIXTURE21 analysis performed at K=3 (Supplementary Figs 2 and 3), which closely matches the Africa, Europe and Asia/Native Americans partition. Continental ancestry estimates are highly correlated (P value <10−12) between all three approaches (Supplementary Fig. 2), although the squared distance between the masked continental ancestry estimates and that estimated by ADMIXTURE21 was, respectively, 5.4-fold and 7.9-fold reduced by the masking procedure for Europe and Asia/Native Americans, suggesting a slight gain in accuracy using this procedure. No major difference is seen for African contributions, while identified donor populations contributing to the mixture were very similar in both approaches; therefore, we henceforth report results on the basis of the masking procedure (Fig. 3).

Bottom Line: The movement of people into the Americas has brought different populations into contact, and contemporary American genomes are the product of a range of complex admixture events.Here we apply a haplotype-based ancestry identification approach to a large set of genome-wide SNP data from a variety of American, European and African populations to determine the contributions of different ancestral populations to the Americas.Our results provide a fine-scale characterization of the source populations, identify a series of novel, previously unreported contributions from Africa and Europe and highlight geohistorical structure in the ancestry of American admixed populations.

View Article: PubMed Central - PubMed

Affiliation: 1] Institute of Legal Medicine, Catholic University, Largo F. Vito 1, Rome 00168, Italy [2] Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UK.

ABSTRACT
The movement of people into the Americas has brought different populations into contact, and contemporary American genomes are the product of a range of complex admixture events. Here we apply a haplotype-based ancestry identification approach to a large set of genome-wide SNP data from a variety of American, European and African populations to determine the contributions of different ancestral populations to the Americas. Our results provide a fine-scale characterization of the source populations, identify a series of novel, previously unreported contributions from Africa and Europe and highlight geohistorical structure in the ancestry of American admixed populations.

No MeSH data available.