Abstract
Admixed populations are formed by the merging of two or more ancestral populations, and the ancestry of each locus in an admixed genome derives from either source. Consider a simple "pulse" admixture model, where populations A and B merged $t$ generations ago without subsequent gene flow. We derive the distribution of the proportion of an admixed chromosome that has A (or B) ancestry, as a function of the chromosome length $L$, $t$, and the initial contribution of the A source, $m$. We demonstrate that these results can be used for inference of the admixture parameters. For more complex admixture models, we derive an expression in Laplace space for the distribution of ancestry proportions that depends on having the distribution of the lengths of segments of each ancestry. We obtain explicit results for the special case of a "two-wave" admixture model, where population A contributed additional migrants in one of the generations between the present and the initial admixture event. Specifically, we derive formulas for the distribution of A and B segment lengths and numerical results for the distribution of ancestry proportions. We show that for recent admixture, data generated under a two-wave model can hardly be distinguished from that generated under a pulse model.
Introduction
Admixture—the interbreeding of previously isolated populations—leaves a distinct signature in the genomes of descendants. These genomes are mosaics of segments from different ancestral origins. By analyzing the lengths and proportions of these segments, population geneticists can reconstruct historical migration events, specifically estimating the timing ($t$) and magnitude ($m$) of population mixing.
Traditional approaches often rely on local ancestry inference, which attempts to map the specific ancestral origin of every genomic locus. However, this process is computationally demanding and prone to error, especially when source populations are genetically similar. This study proposes a more robust alternative: utilizing the distribution of global ancestry proportions (the total fraction of a chromosome from a given source) to infer historical parameters.
Methodology
The researchers developed a mathematical framework based on stochastic modeling to describe ancestry patterns under different historical scenarios.
The Pulse Model
The simplest scenario is a single "pulse" of admixture occurring $t$ generations ago. The authors modeled the ancestry along a chromosome as a two-state renewal process. Using renewal theory, they derived an explicit probability density function (PDF) for the ancestry proportion. This derivation accounts for the fragmentation of ancestral segments due to recombination over time.
The Two-Wave Model
For more complex histories, such as an initial mixing followed by a subsequent wave of migration, the authors employed Laplace transforms. This approach allows for the derivation of segment length distributions and ancestry proportions for nearly any arbitrary migration history. In the specific case of a "two-wave" model, they obtained explicit formulas for the distribution of ancestral segment lengths.
Validation
The theoretical models were validated using Markovian Wright-Fisher simulations to generate "true" ancestry patterns. Additionally, coalescent simulations were used to test the method's robustness when ancestry is estimated from genomic data using standard software like ADMIXTURE.
Results and Discussion
The study demonstrates that the overall distribution of ancestry proportions across chromosomes provides sufficient signal to recover admixture history accurately.
Inference of Admixture Parameters
The authors developed a Maximum Likelihood Estimation (MLE) framework that can be applied to unphased genomic data. As shown in Figure 1, the inferred admixture time ($\hat{t}$) closely follows the true time ($t$). While the variance of the estimates increases for older events as ancestral segments become shorter, the method remains highly effective for recent historical periods.
The Pulse Approximation
A key insight is that for recent admixture events (typically within the last 10 generations), a complex two-wave migration history is often mathematically indistinguishable from a single pulse. As illustrated in Figure 2, the theoretical segment length distributions for a two-wave model match simulation data perfectly. Furthermore, a "best-fit" pulse model (represented by dashed lines) provides a very close approximation to the two-wave data, suggesting that simple models are often sufficient for describing recent admixture.
Conclusion
This paper provides a rigorous mathematical toolkit for analyzing admixed genomes. By deriving explicit distributions for ancestry proportions, the authors offer a method for historical inference that is both accurate and computationally efficient. While the current models assume large population sizes and distinct source populations, they lay the groundwork for future extensions into continuous migration models and the effects of genetic drift.
Summary updated