August 29, 2013

Nuclear sequences of mitochondrial origin and gene flow in Pleistocene Africa

I haven't read this, but the idea of looking at looking at nuclear sequences of mitochondrial origin (numt) as a way of testing for archaic admixture seems interesting to me. Human mtDNA has a relatively shallow coalescence (less than 200 thousand years), with Neandertal mtDNA being a clear outgroup, and Denisovan mtDNA being an even more remote outgroup. If modern humans admixed with archaic ones (having mtDNA lineages much more remote than "Eve"), then the evidence may have been lost (due to drift) in mtDNA, but may have been preserved in the autosomes.

Anthropologischer Anzeiger, Volume 70, Number 2, July 2013 , pp. 221-227(7)

Hominin evolution and gene flow in the Pleistocene Africa

Ovchinnikov, Igor V.

Africa demonstrates a complex process of the hominin evolution with a series of adaptive radiations during several millions of years that led to diverse morphological forms. Recently, Hammer et al. (2011) and Harvati et al. (2011) provided integrated morphological and genetic evidence of interbreeding between modern humans and unknown archaic hominins in Africa as recently as 35,000 years ago. However, a genetic evidence of hybridization between hominin lineages during the Lower and Middle Pleistocene epochs is unknown and the direct retrieval of DNA from extinct lineages of African hominins remains elusive. The availability of both nuclear and mitochondrial genome sequences from modern humans, Neanderthals, and Denisovans allows collecting nuclear DNA sequences of mitochondrial origin (numts) inserted into the nuclear genome of the ancestral hominin lineages and drawing conclusions about the hominin evolution in the remote past. The mtDNA and numt analysis uncovered a deep division of mtDNA lineages that existed in African hominins in the Middle Pleistocene. The first cluster included the human and Neanderthal-like mtDNA sequences while the second consisted of DNA sequences that are known today as mtAncestor-1, a nuclear fossil of the mtDNA, and the Denisova mtDNA isolated from a bone and a tooth found in southern Siberia. The two groups initially diverged 610,000-1,110,000 years ago. Approximately 220,000 years after the primary split, the Denisova - mtAncestor-1 mtDNA lineages mixed with the mtDNA pool of an ancestral population of Neanderthals and modern humans. This admixture after the profound division is demonstrated by the transposition of the Denisova-like mtDNA sequence into the nuclear genome of an ancestor of Neanderthals and modern humans. This finding suggests the matrilineal genetic structure among the Middle Pleistocene hominins as well as the existence of gene flow between African hominin lineages. Through paleogenomic analyses, it is impossible to exclude the theory that population structure and gene flow in African hominins influenced the admixture pattern observed in the nuclear genomes of non-Africans.



Rokus said...

I found the same close relationship with Denisovans like the mtAncestor-1 insert in chromosome 1 for the mtDNA insertion in chromosome 11. Actually, I wonder both inserts are indeed phylogenetically closer to Denisovan mtDNA, or rather indicative of ancestral mtDNA that was slightly different from the one commonly deduced using outgroups. Such mtDNA as a reference would naturally imply a much more complicated history of mutations and back mutations for modern mtDNA.

eurologist said...

Someone wanted $50 for this rather questionable paper for an educational institution access heavily invested in the top 1,000+ publications - for free, for members.

Sorry, ain't going to happen.

Next step is active outcasting of such looser for-profit group.

No scientists should choose such media.

Grognard said...

They choose it for a couple reasons, mostly bad. The biggest one is to silence critics. If they publish in a journal it has fixed reviewers and they all tend to be stacked towards a set of beliefs, so they can get review of obvious mistakes in methodology and data without getting challenged on the actual conclusions, which is usually what's questionable. The second is they often plan to write books and feel if they publish the results publicly they won't get a book deal.

And if you don't publish a book you don't get tenure. But if everyone played by the same rules then you would not have to have it so everyone needed to publish a book. I mean, that is the problem, not the solution to anything. Doing science in such a way as it's going to sell the most books has led to a lot of the junk science we look at every day.

Rokus said...

Please also read Genome Digging: Insight into the Mitochondrial Genome of Homo
by Ovchinnikov & Kholina, 2010.
No comparison with Denisova was yet available, though the scientific world "should" already have picked up the divergence of African mtDNA compared to especially the omnipresent mtAncestor-1 (numt-1) - even in relation with non-African mtDNA. Actually, it is hard NOT to see in table 1 some intermediate position of non-African mtDNA between Neanderthal and African mtDNA for the higher scores. Lower scores identify older inclusions, where differences with Neanderthal, non-African and African mtDNA are "naturally" levelled off. However, the highest scores together (resp. in numt-1, numt-4 and numt-6) apparently still indicate a closer relationship of Neanderthal mtDNA with ancestral mtDNA in the autosomes, and a more distant relationship with African mtDNA, but - strikingly - an intermediate relationship with non-African mtDNA. This should imply a fundamental confict with the current mtDNA phylogenitic tree, and actually should rather reverse the tree, with African diversity the result of more recent branching and the root closer to eg. mtDNA U, X and B.
At least numt-1 (mtAncestor-1) is omnipresent among modern humans, IMO poor evidence of hybridization. A different picture can be discerned for numt-5, the insert of chromosome 11, that is prevalent in non-African populations but not in African populations, and moreover has Neanderthal and modern humans as virtually equivalent outgroups, and is still vaguely reminiscent of Denisovan mtDNA. Now, that is a clear indication of hybridization events in the past! I remain unconvinced the same applies to mtAncestor-1.