Dr. Reich worked with the Max Planck Institute as a population geneticist studying Neanderthal genetic data. This section will discuss the background of his research as part of the Neanderthal genome project, the draft sequence that they assembled, and the evidence that has been compiled for gene flow between modern humans and Neanderthals.
Neanderthals are the only other hominid with a brain as large as Homo sapiens. Neanderthal fossils from 200,000 years ago have been found in West Eurasia (Europe and Western Asia), which is far earlier than Homo erectus. The earliest human fossils come from Ethiopia dating about 200,000 years ago. However, there is evidence that Neanderthals and humans overlapped in time and space between 135,000 and 35,000 years ago.
The first place of contact could have occurred in The Levant, in Israel. There are human fossils from 120,000 years ago, then a gap, Neanderthal fossils about 80,000 years ago, another gap, and then human fossils again 60,000 years ago. This is proof of an overlap in place, but not in time. In the upper paleolithic era, there was an explosion of populations leaving Africa (the migration about 60,000 to 45,000 years ago). In Europe after 45,000 years ago, there are sites where Neanderthals and humans exist side by side in the fossil record. Since there is evidence that the two species co-existed, was there interbreeding? This is a question that can be answered by examining population genomics.
See Tools and Techniques for a discussion of DNA extraction from Neanderthals.
28.5.2 Evidence of Gene Flow between Humans and Neanderthals
- A comparison test between Neanderthal DNA and human DNA from African and non-African populations demonstrates that non-African populations are more related to Neanderthals than African populations. We can look at all the SNPs in the genome and see whether the human SNP from one population matches the Neanderthal SNP. When different human populations were compared to Neanderthals, it was found that French, Chinese, and New Guinea SNPs matched Neanderthal SNPs much more than Nigerian Yoruba SNPs matched Neanderthal SNPs. San Bushmen and Yoruba populations from Africa, despite being very distinct genetically, both had the same distance from Neanderthal DNA. This evidence suggests that human populations migrating from Africa interbred with Neanderthals.
- A long-range haplotype study demonstrates that when the deepest branch of a haplotype tree was in non-African populations, the regions frequently matched Neanderthal DNA. African populations today are the most diverse populations in the world. When humans migrated out of Africa, diversity decreased due to the founder e↵ect. From this history, one would expect that if you built a tree of relations, the deepest split would be African. To show Neanderthal heritage, Berkley researchers picked long range sections of the genome and compared them among randomly chosen humans from various populations. The deepest branch of the tree constructed from that haplotype is almost always from the African population. However, occasionally non-Africans have the deepest branch. The study found that there were 12 regions where non-Africans have the deepest branch. When this data was used to analyze the Neanderthal genome, it was found that 10 out of 12 of these regions in non-Africans matched Neanderthals more than the matched the human reference sequence (a compilation of sequences from various populations). This is evidence of that haplotype actually being of Neanderthal origin.
3. Lastly, there is a bigger divergence than expected among humans. The average split between a Ne- anderthal and a human is about 800,000 years. The typical divergence between two humans is about 500,000 years. When looking at African and non-African sequences, regions of low divergence emerged in non-African sequences when compared with Neanderthal material. The regions found were highly enriched for Neanderthal material (94% Neanderthal), which would increase the average divergence between humans (as the standard Neanderthal - human divergence is about 800,000 years).
Gene Flow between Humans and Denisovans
In 2010, scientists discovered a 50,000 year old finger bone in southern Siberia. The DNA in this Denisovan sample was not like any previous human DNA. Denisovan mitochondrial DNA is an out-group to both Neanderthals and modern humans. (Mitochondrial DNA was used because it is about 1000 times more frequent than somatic DNA. The polymorphism rate is also 10 times higher.) Denisovans are more closely related to Neanderthals than humans.
Using the same SNP matching technique from the Neanderthal example, it was discovered that Deniso- van DNA matches New Guinean DNA more than Chinese DNA or European DNA. It is estimated that Denisovans contributed about 5% of the ancestry of New Guineans today. A princple component analysis projection (see figure) between relatedness to chimpanzees, Neanderthals, and Denisovans shows that non- African populations are more related to Neanderthals, and New Guinean/Bougainvillians are more related to Denisovans.
This evidence suggests a model for human migration and interbreeding. Humans migrated out of Africa and interbred with Neanderthals, then spread across Asia and interbred with Denisovans in Southeast Asia. It is less plausible that humans interbred with Denisovans in India because not all of the populations in Southeast Asia have Denisovan ancestry.
Analysis of High Coverage Archaic Genomes
High-coverage archaic genomes can tell us a lot about the history of hominid populations. A high coverage Altai Neanderthal sequence was acquired from a toe bone found in Denisova cave. From this sequence, we can look at the time to convergence of the two copies of chromosomes to estimate the size of the population. Neanderthal DNA contains many long stretches of homozygosity, indicating a persistant small population size and inbreeding. For the Altai Neanderthal, one eighth of the genome was homozygous, about the expected level of inbreeding of half-siblings. Applying the technique to non-African populations shows a bottleneck 50,000 years ago and a subsequent population expansion, which is consistent with the Out Of Africa theory.
Neanderthals and Denisovans also interbred, demonstrating the remarkable proclivity of humanoids to- wards reproduction. Although most of the Neanderthal genome has a minimum depth of hundreds of thousands of years from the Denisovan genome, at least 0.5% of the Denisovan genome has a much shorter distance from Neanderthal genome, especially for immune genes.
Denisovans most likely have ancestry from an unknown archaic population unrelated to Neanderthals. An African sequence has a 23% match with Neanderthal DNA and 47% match with Denisovan DNA, which is statistically significant. If you stratify the D-statistic by the frequency of an allele in the population, you see an increasing slope and a sharp jump when you reach fixation which most closely matches the predictions one would obtain from an unknown population flowing into Denisovans (see figure).
The bottleneck caused by the migration from Africa is only one example of many. Most scientists usually concentrate on the age and intensity of migration events and not necessarily the duration, but the duration is very important because long bottlenecks create a smaller range of diversity. One way to predict the length of a bottleneck is to determine if any new variations arose during it, which is more likely during longer bottlenecks. The change in the range of diversity is also what helped create the different human sub-populations that became geographically isolated. This is just another way that population genomics can be useful for helping to piece together historical migrations.
Genetic differences between species (here within primates) can be used to help understand the phylogenetic tree from which we are all derived. We looked at the case study of comparisons with Neanderthal DNA, learned about how ancient DNA samples are obtained, how sequences are found and interpreted, and how that evidence shows high likelihood of interbreeding between modern humans (of Eurasian descent) and Neanderthals. Those very small differences between one species and the next, and within species, allow us to deduce a great deal of human history through population genetics.