NEW YORK (GenomeWeb Daily News) – Drawing on regions of the human genome that are under neutral selective pressures, researchers led by Cornell University's Alon Keinan found that the human population has increased by about 100-fold during the past 3,000 to 4,000 years.
As they reported today in the Proceedings of the National Academy of Sciences, Keinan and his colleagues focused on loci in the human genome that contained neutral mutations and sequenced them to high coverage in 500 people with homogenous northwestern European ancestry. With this data, the researchers developed several models of recent European demographic history, and the best-fit model indicated about a 3.4 percent population growth rate per generation for the last 140 generations or so.
Modern humans have undergone recent rapid population growth, which is thought to have been triggered by the Neolithic revolution about 10,000 years ago, the researchers said. Population growth has continued and even increased in the last 2,000 or so years to reach the current population of more than 7 billion people.
Previous estimates have placed the increase of effective population size at about 0.5 percent per generation. Such studies, Keinan and his colleagues noted, provided estimates that relied on protein-coding genes, which are affected by natural selection. But in their study, he and his colleagues aimed to minimize the effects of both positive and purifying selection on the gene variants used to gauge population growth as well as to control for the effects of population structure.
"[W]e aim to capture recent demographic history and estimate the magnitude of the recent growth experienced by humans while limiting the confounding by natural selection and population structure," the researchers said.
To do so, they developed a set of neutral regions that contain loci at least 100,000 basepairs or 0.01 centimorgans away from potentially coding regions, that do not contain known CNVs or segmental duplications, and that do not contain a number of highly conserved or repetitive regions. They sequenced 15 loci totaling 216,240 basepairs that met these criteria.
To find a sample with similar ancestry, which would minimize biased effects of rare variants or singletons, the researchers conducted a principal components analysis of whole-genome genotyped data of 9,716 people of European descent from the Atherosclerosis Risk in Communities dataset. The researchers then based their analyses on a group of 500 people that formed a homogenous cluster. Validation using the POPRES dataset indicated that this subsample was less heterogeneous than the ARIC sample as a whole.
The researchers sequenced those 15 loci in the set of 500 people to very high coverage using the Illumina HiSeq platform and got a median average depth of coverage of 295x per person. After applying the UnifiedGenotype tool to detect variants and make genotype calls, and after filtering those calls, they wound up with 1,834 high-quality SNVs, which they validated using whole-genome sequencing data from the Cohorts for Heart and Aging Research in Genetic Epidemiology project as well as with data from the Exome Sequencing Project.
Of those SNVs, 62.5 percent had not previously been reported in dbSNP, the researchers noted.
Keinan and his colleagues developed a number of models of human population growth that they then compared to the SNV data.
The best-fit model indicates that the European population experienced recent growth from a population size of about 4,000 people to 7,000 people some 120 generations to 160 generations ago, or about 3,000 years to 4,000 years ago. That growth proceeded at about 2 percent to 5 percent per generation, leading to an increase in effective population size of two orders of magnitude. This, they noted, is faster than previous estimates had indicated.
The best-fit model included three free parameters: the time at which growth started, extant effective population size, and effective population size just before exponential growth.
Keinan and his colleagues noted, though, that assumptions regarding ancient demography influence estimations of recent population growth. "We hypothesize that this is the case because previous models of ancient demographic history resulted in parameters that confound more recent and more ancient history, with the recent growth indirectly affecting them in a manner dependent on sample size," they said. "This realization leads to the model we report here fitting much better than previous models of recent growth, and it sheds light on the discrepancies among the latter."
They also examined models that included an acceleration of the growth rate, as archaeological evidence has indicated the growth began with the Neolithic revolution, but none of those models fit the genetic data.
"These models can inform studies of natural selection, the architecture of complex diseases, and the methods that should best be used for genotype-phenotype mapping," Keinan and his colleagues noted.