Comment by 08-15 | Hacker News Reader

Comment by 08-15 | Hacker News Reader

Parent Full thread

It is far more notable that there is exactly one known Denisovan fossil, namely the distal phalanx the whole "Denisovan Genome" was sequenced from. One could argue that she was just another Neanderthal, but that didn't do the vanity of the authors of "Genetic history of an archaic hominin group from Denisova Cave in Siberia" [Reich2010] justice.

Regarding the admixture, the analysis doesn't indicate the direction. If a Denisovan had offspring with an ancestor of present day Philippinos and the offspring stayed with the humans, that would create exactly the same D-statistics as if the child stayed with the Denisovans.

I think the most believable scenario is that humans migrated through an area that was already inhabited by Neanderthals, some of them had fun with the locals, and the Denisovan is an offspring of such a family. That's much easier to believe than Denisovans, who somehow never left traces in the fossil record, living all over Asia and contributing genes to all present day Melanesians.

[Reich2010] has a single sentence to rule out this scenario: the fossil is a tiny bit too old to be affected by the migrating modern humans. That relies on stratigraphic dating (C14 fails at around 50ka), and that relies on the assurance of the archaeologists that "most of the site was disturbed, but this fossil is from the undisturbed part!" Yeah, right.

As you said, Denisovans, assuming there was ever more than one, and assuming they weren't just Neanderthals, probably swam to the Philippines. Isn't it amazing how much we can learn from a single pinky bone?

But there are multiple known Denisovans by now, i.e https://www.pnas.org/content/112/51/15696

There is even a known Neanderthal Denisovan hybrid: https://www.nature.com/articles/d41586-018-06004-0

(Of course, that doesn't make it less amazing how much we can learn from a single bone)

Well, there are two more Denisovans according to their mitochondrial genome. There isn't enough nuclear genomic data from either to measure the distance to Neanderthals or confirm the alleged admixture with Melanesians. Too bad, because if it wasn't there, it would confirm my conjecture that the flow was from modern humans to Neanderthals.

The "hybrid" is just more confused terminology. If we simply acknowledged, that Neanderthals had higher genetic diversity than previously thought, the Denisovan and the so-called hybrid would just be a Neanderthal. We don't talk about hybrids of Europeans and Africans either, do we? If I remember correctly, the admixture wasn't confirmed in the hybrid, either. I think, nobody looked.

The root of this nonsense is that Neanderthals were thought to be genetically very similar, especially their mitochondria. The reason is that prior to 2009, it wasn't possible to routine sequence the mitochondrion, let alone the nuclear genome, of some random fossil. Therefore, all fossils were pre-screened using a PCR test for a mitochondrial mutation specific to Neanderthals. But not all Neanderthals have that mutation, so those that were sequenced were very similar due to a technical limitation.

We have plenty of Neanderthal genetic sequences. So seeing even just one Denisovan sequence showing a large divergence from all the other Neanderthal sequences is enough to demonstrate a separate, distinct population.

How much is large?

The dendrogram is right here: https://www.eva.mpg.de/genetics/genome-projects/neandertal/

The difference between the Denisovan and the Neanderthals is bigger than between San and French, but not by much. Keep in mind that the archaic genomes are noisy, and noise adds to their distance.

At the time, the debate was whether Neanderthals and the Denisovan constitute different species or just different subspecies. (Taxonomy is fun. You get to name species you discover after yourselves. I'd love to discover a new species, but in a pinch, a subspecies will do.)

You use the meaningless term "population". Whatever, let them be different populations. Now we have two populations, namely the "Denisovan" and the "Altai Neanderthal" living in the same cave, at roughly the same time.

San and French lines have been separated by 50-100k years at least, right? So if Denisovans are more different from Neanderthals than that, why do you think the Denisovan bone was from a Modern x Neanderthal instead?

Also, if you did find a first generation Modern x Neanderthal, you'd have stretches of the genome that were 100% Neanderthal and stretches that were 100% Modern. It'd be obvious that it was an F1 hybrid. Even F2, F3 etc it'd be pretty clear. Are you saying the Denisovan genome doesn't have unique mutations?

I agree that the terms species / subspecies / populations are difficult/impossible to to define sharply.

No, not first generation. That would indeed show up, as it did in "Oase 1": https://www.nature.com/articles/nature14558

> Are you saying the Denisovan genome doesn't have unique mutations?

It sure does, but we don't know them. Anything unique to a single archaic individual can always be a sequencing error. (But it's high coverage! Right, but they couldn't compute proper genotype likelihoods for ancient DNA, so systematic errors end up in the final sequence.) And anything found only in Denisovans, however you define them, may or may not be fixed in them, and may or may not exist in other hominins. With so few archaic genomes, it's impossible to tell.

The actual analysis is done on variants that exist in extant humans. Which is why you never read sentence about the number of mutations unique to whoever, you always get something like "is more similar to X than to Y".

> So if Denisovans are more different from Neanderthals than that, why do you think the Denisovan bone was from a Modern x Neanderthal instead?

Distance is calculated by walking along the genome, calling a genotype for every base, and counting difference. This works well for modern humans, but for Neanderthals, a good chunk of the called genotypes are erroneous. This increases their apparent distance from humans, but it increases their apparent distance from each other even more. I'm saying, the Denisovan and Neanderthals may not really be more different from each other than French and San.

I'm not saying anything about the Denisovan in particular. I'm saying the gene flow was probably into Neanderthals, and we stumbled upon one affected individual. Interbreeding is a local event. It is easy to believe that we stumbled upon an individual that had some human ancestry. It is much harder to believe that such an individual ended up in a human population and ended up affecting the whole population uniformly.

Denisovans, Neanderthals, and sapiens could all interbreed with each other without issue, and only diverged about half a million years ago, so it is only reasonable to say we’re all just different subspecies.

Sorry, minor point: the country is the Philippines, the language and people are Filipino.

Reply on news.ycombinator.com