Neanderthal Genome Project

In July 2006, the Max Planck Institute for Evolutionary Anthropology and 454 Life Sciences announced that they would be sequencing the Neanderthal genome over the next two years. At three billion base pairs, the Neanderthal genome is roughly the size of the human genome; preliminary sequences reveal very few differences: human and Neanderthal DNA appear to be 99.5% percent identical.

The researchers recovered Neanderthal DNA by extracting all the DNA from the femur bone of a 38,000-year-old male Neanderthal specimen from Vindija Cave, Croatia.

Results of two groups working on the same Neanderthal sample have been published in Nature^[1] and in Science.^[2] This results have been received with a fiery polemic, the core issue of which was captured by this comment: "Although there won't be any Neanderthals around to appreciate their legacy being uncovered, it's possible that these results will help us find a little bit of Neanderthal in all of us."^[3]

Additional recommended knowledge

Daily Sensitivity Test

Weighing the right way

Don't let static charges disrupt your weighing accuracy

Project Results

The group of Green et al used a new sequencing technique that amplifies single molecules for characterization and obtained over a quarter-million unique sequences. The group of Noonan et al used different technique, one in which the Neanderthal DNA is inserted into bacteria, which make multiple copies of a single fragment. It's a slower technique and only 65,000 bases were sequenced, but the same DNA can be obtained from the bacteria as needed and thus allows for a higher degree of error correction. Overall, their results were remarkably similar. One group suggested there was a hint of mixing between human and Neanderthal genomes, while the other found none, but both recognize that the data set is just not large enough to give a definitive answer.^[3]

Within the project, Svante Pääbo directly sequence the Neanderthal nuclear DNA genome. Since direct sequencing is random, one must wait for specific sequences for genes that might be different between modern humans and Neanderthals to show in the process. Direct sequencing destroys the original sample, so in principle the metagenomic library approach will forever retain a clone of the Neanderthal DNA for future targeted research.

Importantly, the project researchers led by Edward Rubin also demonstrated that Neanderthal genomic sequences can be recovered using a metagenomic library-based approach. All of the DNA in the sample is "immortalized" into metagenomic libraries. A DNA fragment is selected, then propagated in microbes. The Neanderthal DNA can be sequenced or specific sequences can be studied.^[4]

Earlier mitochondrial DNA research led by Svante Pääbo in 1997 indicated archaic Homo sapiens and Neanderthals broke into separate lineages approximately 500,000 years ago. The publication in Science, 2006, revealed Neanderthal DNA sequences matching chimpanzee DNA but not modern human DNA at multiple locations, thus enabling the first accurate calculation of the date of the most recent common ancestor of H. sapiens and H. neanderthalensis. This nuclear DNA study indicates that the common ancestor lived about 706,000 years ago, and that the modern European-Neanderthal population split occurred 330,000 years later.^[4] However, Green et al. calculated a divergence time of 516 kya and don't indicate a split.

Peer Review

The peer review of Wall and Kim^[5] reanalyses the data obtained from the published papers of Noonan et al.^[2] and Green et al.,^[1] and holds the results not consistent one with the other. The review proposes serious problems with the data quality in one of the studies, possibly due to modern human DNA contaminants and/or a high rate of sequencing errors.

The reanalyses conveniently confirmed both results to the Human-Neanderthal DNA Sequence Divergence Time (common ancestor), that is 706 kya to the Noonan et al. analysis and 516 kya to the Green et al. analysis. The modern European-Neanderthal population split time was estimated 35 kya for the Green et al. data, and 325 kya for the Noonan et al. data. Before, no split time was estimated by the Green et al. study, and according to Wall and Kim the split time originally estimated by Noonan et al. was even higher: 440 kya (the Noonan et all paper mentions 370 kya^[1]).

While Noonan et al. were unable to definitively conclude that interbreeding between the two species of humans did not occur, they proclaim the low likelihood of it having occurred at any appreciable level.^[4] The study opt to a 0% contribution of Neanderthal DNA to the modern European gene pool, based on the 95% confidence interval that indicates a margen between 0% and 20% contribution. The reanalyses of Wall and Kim yielded interbreeding margens between 0% and 39% to the data of Noonan et al., and margens between 81% and 100% to the data of Green at al. This vastly inconsistent results could only be reconciled by assuming a very recent split time between the two populations of 60 kya or less. However, such a recent split time would not be consistent with the estimated modern European-Neanderthal population split time from the Noonan et al. data.

The key assumption to Noonan et al. ^[1][6] is the 38,000 years of fossilisation that the Neanderthal DNA suffered should have the genome analysis focus on ancient DNA fragments of about 50 to 70 base pairs in length. Green et al. does not make such an assumption and generalized towards the exclusion of modern human nuclear DNA contamination by finding little evidence of modern human mtDNA contamination. Such mitochondrial DNA tends to remain preserved longer than nuclear DNA.^[4] However, Wall and Kim noted a length dependence of the results, having the small fragments pointing to a divergence time similar to the results of Noonan et al. and the large fragments much more similar on average to modern human DNA: even to the extend of indicating at an estimated human-Neanderthal sequence divergence time that is less than the estimated divergence time of two extant members of one referenced population in West Africa. Though Wall and Kim hold modern human contamination to be size biased, since actual Neanderthal DNA would be expected to have a tendency to be degraded into short fragments, they noted that the observation of a length dependence of the results makes alignment issues alone unlikely to be a sufficient explanation, since longer fragments would be easier to align and thus the data from longer fragments should be more accurate. Still they mark this as a signal of potential contamination in the data of Green et al. No similar signal of potential contamination was found in the data of Noonan et al.

Contamination in the data of Green et al. should have decreased the Neanderthal-specific sequence divergence to this study. Since this is not the case, the assumption of contamination would also indicate a higher sequencing error rate in the Green et al. data, since sequence errors would look the same as Neanderthal-specific mutations. It should be noted this Neanderthal-specific mutations were already considered prone to error due to post-mortem DNA damage in both studies, and were excluded from the results. ^[7]

In summary, Wall and Kim consider a model with 78% contamination more likely than a model with no contamination and 94% admixture.

Notes

See also

• Neanderthal
• Neanderthal interaction with Cro-Magnons