NEW YORK (GenomeWeb News) – African and non-African populations continued swapping genetic information long after the main wave of human migration out of Africa some 60,000 years ago, according to a study by a pair of Wellcome Trust Sanger Institute researchers appearing online in Nature today.
The team used a new statistical method to track human population patterns using information on heterozygous alleles across the diploid genomes of sequenced individuals. Using published individual genome sequence data for seven men — one from China, another from Korea, two from the Yoruban population in West Africa, and three of European ancestry — the researchers gained insights into the human population sizes and bottlenecks that occurred between 10,000 years and one million years ago.
In contrast to some past studies, they found continued genetic exchange between African and non-African populations from the time that the populations began to be distinguishable from one another around 100,000 years ago until long after the main wave of migration out of Africa some 60,000 years ago.
"There was genetic exchange between the African ancestral population and the European ancestral population throughout this period, from 100,000 years ago right through to what our model is suggesting is 20,000 years ago," senior author Richard Durbin, genome informatics leader at the Sanger Institute and joint head of the institute's human genetics program, told GenomeWeb Daily News.
Previous strategies for gauging shared common ancestry have largely relied on information from just a few spots in the nuclear genome, such as sites on the Y chromosome, or from mitochondrial genome data, Durbin explained.
"You can look at lots of individuals and you can build a tree that relates the individuals and finds the most recent common ancestor of the people in that tree," he explained. "The problem is that you're only looking at a single locus — it's one possible common ancestor for that particular genomic element."
For the current study, Durbin and co-author Heng Li came up with a statistical approach, called a pairwise sequentially Markovian coalescent model, which uses information from the entire diploid genome.
By focusing on the density and distribution of heterozygous sites at which two copies of an allele differ from one another, the duo was able to account for local areas of recombination and estimate the time since heterozygous alleles shared the most recent ancestor.
"Each human genome contains information from the mother and the father, and the differences between these at any place in the genome carry information about its history," Li explained in a statement.
"Since the genome sequence is so large, we can combine the information from tens of thousands of different places in the genome to build up a composite history of the ancestral contributions to the particular individual who was sequenced," Li added. "We can also get at the historical relationship between two different ancestral populations by comparing the X chromosomes from two males."
The duo tested their approach using autosome and X chromosome data from published genome sequences for individuals from West Africa, Korea, China, and for individuals of European descent.
Their data suggest that West African and non-African populations belonged to the same ancestral population some 100,000 years ago, though it was possible to begin distinguishing the populations from one another genetically around that time. This separation, while incomplete, hints that there was some population differentiation that occurred prior to the main out-of-Africa migration around 60,000 years ago, Durbin explained.
Human populations apparently became relatively large slightly before that time, the researchers reported, coinciding with the advent of modern humans. "It looks as though, around the time of modern humans arising between 100,000 and 200,000 years ago or so, there was an increase in effective population size," Durbin said.
These populations subsequently experienced bottlenecks leading to reduced population sizes, though the extent of these bottlenecks varied between the African and non-African populations.
The researchers estimate that the European and East Asian populations experienced steep bottlenecks some 10,000 to 60,000 years ago, coinciding with the out-of-African migration, while the African population tested experienced a milder bottleneck from which it recovered more quickly.
The new data also points to widespread genetic exchange between African and non-African populations as recently as 20,000 to 40,000 years ago — long after a prominent wave of migration from Africa around 60,000 years ago.
"The hypothesis that there was significant ongoing genetic exchange throughout the [out-of-Africa] bottleneck is surprising in light of current views about human migrations," Li and Durbin wrote, "however, it is not inconsistent with the archaeological literature, and should motivate further research."
The pair conceded that there is some degree of uncertainty about the per-year mutation rate and generation times used in the model. And, they say, it's possible that there were unknown sex biases in human demographic patterns that might alter the interpretation of X chromosome data.
Still, they say, the method relies on fewer assumptions overall than approaches used in the past — and holds potential for studying other species with sequenced diploid genomes. Down the road, a similar strategy may also be useful for looking at larger human sample, Durbin said, though he emphasized that there are technical hurdles to overcome before that happens.
"If we looked at more people, we could see more recent events, but that would complicate the analysis," he explained. "I think it is likely that there will be such models in the future, either from our group or other groups."