A ghost population is a population that has been inferred through using statistical techniques.
In 2004, it was proposed that maximum likelihood or Bayesian approaches that estimate the migration rates and population sizes using coalescent theory can use datasets which contain a population that has no data. This is referred to as a "ghost population". The manipulation allows exploration in the effects of missing populations on the estimation of population sizes and migration rates between two specific populations. The biases of the inferred population parameters depend on the magnitude of the migration rate from the unknown populations. The technique for deriving ghost populations attracted criticism because ghost populations were the result of statistical models, along with their limitations.
In 2012, DNA analysis and statistical techniques were used to infer that a now-extinct human population in northern Eurasia had interbred with both the ancestors of Europeans and a Siberian group that later migrated to the Americas. The group was referred to as a ghost population because they were identified by the echoes that they leave in genomes — not by bones or ancient DNA. In 2013, another study found the remains of a member of this ghost group, fulfilling the earlier prediction that they had existed.
According to a study published in 2020, there are indications that 2% to 19% (or about ≃6.6 and ≃7.0%) of the DNA of four West African populations may have come from an unknown archaic hominin which split from the ancestor of humans and Neanderthals between 360 kya to 1.02 mya. However, the study also suggests that at least part of this archaic admixture is also present in Eurasians/non-Africans, and that the admixture event or events range from 0 to 124 ka B.P, which includes the period before the Out-of-Africa migration and prior to the African/Eurasian split (thus affecting in part the common ancestors of both Africans and Eurasians/non-Africans). Another recent study, which discovered substantial amounts of previously undescribed human genetic variation, also found ancestral genetic variation in Africans that predates modern humans and was lost in most non-Africans.
In 2015, a study of the lineage and early migration of the domestic pig found that the best model that fitted the data included gene flow from a ghost population during the Pleistocene that is now extinct.
- Beerli, P (2004). "Effect of unsampled populations on the estimation of population sizes and migration rates between sampled populations". Molecular Ecology. 13 (4): 827–836. doi:10.1111/j.1365-294x.2004.02101.x. PMID 15012758. S2CID 18326408.
- Skatkin, M (2005). "Seeing ghosts: the effect of unsampled populations on migration rates estimated for sampled populations". Molecular Ecology. 14 (1): 67–73. doi:10.1111/j.1365-294X.2004.02393.x. PMID 15643951. S2CID 17600283.
- Patterson, N (2012). "Ancient admixture in human history". Genetics. 192 (3): 1065–93. doi:10.1534/genetics.112.145037. PMC 3522152. PMID 22960212.
- Raghavan, M (2013). "Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans". Nature. 505 (7481): 87–91. doi:10.1038/nature12736. PMC 4105016. PMID 24256729.
- Callaway, E (2015). ""Ghost population" hints at long-lost migration to the Americas". Nature. doi:10.1038/nature.2015.18029. S2CID 181337948.
- Arun Durvasula; Sriram Sankararaman (2020). "Recovering signals of ghost archaic introgression in African populations". Science Advances. 6 (7): eaax5097. Bibcode:2020SciA....6.5097D. doi:10.1126/sciadv.aax5097. PMC 7015685. PMID 32095519. "Non-African populations (Han Chinese in Beijing and Utah residents with northern and western European ancestry) also show analogous patterns in the CSFS, suggesting that a component of archaic ancestry was shared before the split of African and non-African populations...One interpretation of the recent time of introgression that we document is that archaic forms persisted in Africa until fairly recently. Alternately, the archaic population could have introgressed earlier into a modern human population, which then subsequently interbred with the ancestors of the populations that we have analyzed here. The models that we have explored here are not mutually exclusive, and it is plausible that the history of African populations includes genetic contributions from multiple divergent populations, as evidenced by the large effective population size associated with the introgressing archaic population...Given the uncertainty in our estimates of the time of introgression, we wondered whether jointly analyzing the CSFS from both the CEU (Utah residents with Northern and Western European ancestry) and YRI genomes could provide additional resolution. Under model C, we simulated introgression before and after the split between African and non-African populations and observed qualitative differences between the two models in the high-frequency–derived allele bins of the CSFS in African and non-African populations (fig. S40). Using ABC to jointly fit the high-frequency–derived allele bins of the CSFS in CEU and YRI (defined as greater than 50% frequency), we find that the lower limit on the 95% credible interval of the introgression time is older than the simulated split between CEU and YRI (2800 versus 2155 generations B.P.), indicating that at least part of the archaic lineages seen in the YRI are also shared with the CEU..."
-  Supplementary Materials for Recovering signals of ghost archaic introgression in African populations", section "S8.2" "We simulated data using the same priors in Section S5.2, but computed the spectrum for both YRI [West African Yoruba] and CEU [a population of European origin] . We found that the best fitting parameters were an archaic split time of 27,000 generations ago (95% HPD: 26,000-28,000), admixture fraction of 0.09 (95% HPD: 0.04-0.17), admixture time of 3,000 generations ago (95% HPD: 2,800-3,400), and an effective population size of 19,700 individuals (95% HPD: 19,300-20,200). We find that the lower bound of the admixture time is further back than the simulated split between CEU and YRI (2155 generations ago), providing some evidence in favor of a pre-Out-of-Africa event. This model suggests that many populations outside of Africa should also contain haplotypes from this introgression event, though detection is difficult because many methods use unadmixed outgroups to detect introgressed haplotypes [Browning et al., 2018, Skov et al., 2018, Durvasula and Sankararaman, 2019] (5, 53, 22). It is also possible that some of these haplotypes were lost during the Out-of-Africa bottleneck."
- Bergström, A; McCarthy, S; Hui, R; Almarri, M; Ayub, Q (2020). "Insights into human genetic variation and population history from 929 diverse genomes". Science. 367 (6484): eaay5012. doi:10.1126/science.aay5012. PMC 7115999. PMID 32193295. "An analysis of archaic sequences in modern populations identifies ancestral genetic variation in African populations that likely predates modern humans and has been lost in most non-African populations...We found small amounts of Neanderthal ancestry in West African genomes, most likely reflecting Eurasian admixture. Despite their very low levels or absence of archaic ancestry, African populations share many Neanderthal and Denisovan variants that are absent from Eurasia, reflecting how a larger proportion of the ancestral human variation has been maintained in Africa."
- Frantz, L (2015). "Evidence of long-term gene flow and selection during domestication from analyses of Eurasian wild and domestic pig genomes". Nature Genetics. 47 (10): 1141–1148. doi:10.1038/ng.3394. PMID 26323058. S2CID 205350534.