Did Denisovans really disappear?

[Graeme M]

This question probably shows my ignorance about how genetics work. I have just finished a great book outlining the current state of play regarding our understanding of human evolution and how modern humans gradually spread around the world, displacing other human lineages such as Neanderthals and Denisovans. One oddity is that only modern humans remain - the last of our near relatives disappeared thousands of years ago.

However, their vestiges remain in our genes with segments of Neanderthal and Denisovan in our genetic makeup, resulting from interbreeding events in the distant past. What I am curious about is how we can be confident that we are looking at modern humans with Neanderthal and Denisovan introgression rather than Neanderthals or Denisovans with human introgressian.

That is, when we look at the people of Melanesia for example we learn that they have the most Denisovan DNA at around 4-6% of their genome. This seems a lot. I imagine that as modern humans and Denisovans can interbreed our DNA must be very, very similar. If Denisovans were slowly pushed east by modern humans but did interbreed to some extent, why are we certain that Melanesian folk are modern humans and not Denisovans?

 

[MidgetDwarf]

100%- 6% =?

 

[Graeme M]

100%- 6% =?

That isn't really addressing what I asked. What I am getting at is that modern humans and Denisovans must have very similar DNA. I read that in most African and European populations there is next to zero Denisovan DNA, but it is much higher (4-6%) in Melanesians which presumably is a result of interbreeding in South East Asia/Sahul. I don't know just what that means.

After all, it is often said that we share 95% of DNA with chimpanzees. Presumably these two statements are not talking about the same thing. But I would assume that most of the DNA is the same between a modern person and a Denisovan (95%? 98%? 99%?). So we are only really talking about whatever small proportion of our genome that is different from that of the Denisovans. A better question then in the context of the very small difference in genome between us and a Denisovan might be what does it really mean for a modern human to have inherited 6% of their DNA from a Denisovan?

Regardless, what I am getting at is why we conclude that we are looking at modern humans with some Denisovan ancestry, rather than Denisovans with modern human ancestry. In either case, it seems to me that we should see much the same thing. Most of the genome would match and some small part would be Denisovan. What is it that tells us which group we are looking at?

 

[pinball1970]

I have just finished a great book outlining the current state of play regarding our understanding of human evolution and how modern humans gradually spread around the world, displacing other human lineages such as Neanderthals and Denisovans.

Which book?

 

[russ_watters]

Regardless, what I am getting at is why we conclude that we are looking at modern humans with some Denisovan ancestry, rather than Denisovans with modern human ancestry. In either case, it seems to me that we should see much the same thing. Most of the genome would match and some small part would be Denisovan. What is it that tells us which group we are looking at?

Where the tree branches and which branch we are on. We did not descend from Denisovians, we and Denisovians descended from a common ancestor.

[Though I think our evidence for what the tree looks like at that level of detail is thin.]

 

[Graeme M]

The book is "The World Before Us: How Science is Revealing a New Story of Our Human Origins" by Tom Higham, an English archaeological scientist. I found it very informative.

Just to clarify, I'm not suggesting anything about humans and Denisovans being the same "species". Denisovans branched off long before modern humans. I am more getting at my curiosity about the reasons for classifying someone with 6% Denisovan DNA as human with Denisovan ancestry, rather than Denisovan with human ancestry.

I think about that like this. The only way we can find Denisovan DNA in a modern human's genome is by humans and Denisovans interbreeding and the hybrid offspring continuing to viably breed with other humans. But they must equally have interbred successfully with other Denisovans. Given modern humans and Denisovans must share for the most part almost identical DNA, what is the marker for identifying a modern person with so much Denisovan DNA as modern human rather than modern Denisovan.

 

[russ_watters]

I am more getting at my curiosity about the reasons for classifying someone with 6% Denisovan DNA as human with Denisovan ancestry, rather than Denisovan with human ancestry.

Because 6%<50%? I feel like I must be missing something because this doesn't seem complicated...

But they must equally have interbred successfully with other Denisovans.

Why?

Given modern humans and Denisovans must share for the most part almost identical DNA, what is the marker for identifying a modern person with so much Denisovan DNA as modern human rather than modern Denisovan.

You seem to be implying 6% >> 6%. I think 6% is just 6%.

Is the issue that you think the fraction that is shared is counted with the 6%? That the commonality is really (guessing) 90%+6%=96%? I don't think that's what it means. It's 6% of the part that's different. E.G., 6% of that 10% or 0.6% of the total genes.

 

[BillTre]

There are many definitions for what a species is. It is not clear which you are using or why.

One way to think about it is continuity of identity based on lineage. Being in a species gets inherited. This become problematic when lineages are reticulate or anastomosing, merging together rather than branching apart. The resulting lineage would then inherit stuff, including maybe species identity, from both parents.
Some would call this a hybrid species or a hybrid taxa if not wanting to be so emphatic about the species issue.

This comes down to naming issues, which can be resolve. I usually try to avoid then.
The part of the issue based on percentages of content is somewhat related to The ship of Theseus paradox, where the identity of a ship is questioned based on how many of its pieces are replaced or not.

Percentage of genome wins would be the most common approach.
There are some related issues:
Humans could in this approach, be considered to come in different flavors: human only genome (African derived, African plus Neanderthal, African plus Denisovan, Denisovan plus Neanderthal, and other more complex mixtures. These mixes, composed of hundreds of genetic factors (or markers) would be difficult to cleanly separate. It would probably result an untenable naming systems similar to the racial naming in some anthropological circles during the early 1900's.
I don't know how many researchers would really call humans and Neanderthals different species anyway. Thinking about different breeding populations rather than different species makes a lot of these problems go away. The word species carries a lot of definitional luggage.

Introgression between populations seems to be a pretty close comparison. Two breeding populations meet, breeding ensues along the borders between the populations and genes are passed between the populations while the populations largely maintain their overall character. It is more of an exchange of genetic material than an overall "character" change.

There will be no single marker for being human vs. Neanderthal or Denisovan. Many markers are involved. Not all individuals of a population will inherited the same markers of one population or the other.

 

[Graeme M]

Introgression between populations seems to be a pretty close comparison. Two breeding populations meet, breeding ensues along the borders between the populations and genes are passed between the populations while the populations largely maintain their overall character. It is more of an exchange of genetic material than an overall "character" change.

There will be no single marker for being human vs. Neanderthal or Denisovan. Many markers are involved. Not all individuals of a population will inherited the same markers of one population or the other.

Thanks BillTre. That seems more what I am wondering about. What tells us that Melanesian folk, for example, are modern humans with Denisovan ancestry rather than Denisovans with modern human ancestry? Perhaps russ_watters is explaining that:

Is the issue that you think the fraction that is shared is counted with the 6%? That the commonality is really (guessing) 90%+6%=96%? I don't think that's what it means. It's 6% of the part that's different. E.G., 6% of that 10% or 0.6% of the total genes.

I think this is where I am being confused. What does it really mean to have 6% Denisovan DNA in one's genome? If Denisovans had hung around on some islands somewhere, breeding with hybrids and spreading modern human DNA throughout that population, would the end result look significantly different from the actual situation today?

 

[256bits]

I think this is where I am being confused. What does it really mean to have 6% Denisovan DNA in one's genome? If Denisovans had hung around on some islands somewhere, breeding with hybrids and spreading modern human DNA throughout that population, would the end result look significantly different from the actual situation today?

I don't know what that 6% means.

Perhaps it means that the population interaction between sapiens and Denisovans in ancient times is estimated to have been around 6% in exchange of genetic material. That said I would not think that not everyone of the population living today in Melanesia has an actual 6% of the Denisovan genome. Does everyone in the region have an ancestor who is Denisovan. If they could trace their ancestors, when and of which one would they find of their great grandfather or grandmother as being Denisovan or human?

Mitochondrial DNA is passed from mother to daughter to granddaughter. The mtDNA lineage would come to an abrupt halt if one of the granddaughters down the line has only male offspring. mDNA would be passed on to male offspring though from each parent.

This is a lot more complicated than Mendel and his yellow and green wrinkled peas.

 

[russ_watters]

I think this is where I am being confused. What does it really mean to have 6% Denisovan DNA in one's genome? If Denisovans had hung around on some islands somewhere, breeding with hybrids and spreading modern human DNA throughout that population, would the end result look significantly different from the actual situation today?

I dont know what "significantly different" means. Let's simplify. Denisovian or Scottish, if you have 1 ancestor 4 generations ago that was Scottish that would make you .5^4 = 6% Scottish. If all the rest of your ancestors were Aboriginal Australian what color do you think your skin would be?

 

[BillTre]

Numbers like 6% probably represent some percentage of molecular markers that can be attributed to one or the other group. I am guessing that the percentage is calculated based on the numbers of markers, rather than the percentage of base pairs. Each marker would probably many sequence elements that would mostly be shared among close relatives.

These are numbers generated from sequence surveys.
The numbers of people (individuals) sampled in a population can vary. This affects detection of low frequency sequences that are not prevalent in the population, but still available for evolution.

If there are no selective influences on different sequences (one is not more adaptive than others), then in a randomly breeding population of fairly large size, the fraction of sequences from different sources in a population should not change on average over generations.

To me, this indicates that the 5 or 6% DNA were probably sequence prevalences established in the population after the two populations met and their more deleterious (bad) genes were already eliminated and replaced by their alternatives.

 

[jim mcnamara]

Definition for cladistics (Oxford English Dictionary)
a method of classification of animals and plants according to the proportion of measurable characteristics that they have in common. It is assumed that the higher the proportion of characteristics that two organisms share, the more recently they diverged from a common ancestor.

At some point of convergence you are dealing with populations of the same species. So definitions are important. Domestication messes this up.
Example: dogs

So cladistics based on superficial characteristics can lead investigators astray.

Definition: Taxonomy has used, for years, cladistic data to populate cubbyholes with related species. DNA is a newcomer to the toolset to do this.

Using traditional cladistics water lilies and lotus were assumed to be closely related. DNA says the lotus is the closest living relative to the plane tree (harewood or sycamore tree are other common names), not water lilies.

My point:
This is why it seems a stretch to make Neanderthal, Denisovan, and moderns separate species using either DNA or other cladistics to create 3 distinct and separate taxonomic cubbyholes unless you want to "prove" the superiority of modern humans. Which clearly was evident in early documents on the subject of human evolution. Not today:
Example:
the stooped over posture drawings in expositions circa 1950.

Did you know: modern human brain size is smaller than neolithic human brain size and a few neanderthals as well? Well, maybe.

There is a debate about this, exactly the kind of debate you see in anthropology frequently.
No brain shrinkage:
https://www.unlv.edu/news/release/unlv-research-no-human-brain-did-not-shrink-3000-years-ago
Yes Brain shrinkage:
https://www.frontiersin.org/articles/10.3389/fevo.2021.742639/full

Why did I mention this? Because the topic is anthropology. This kind of debate is what us outsiders may see as hoo-hah - but is really how scientific debate proceeds in that discipline. So do not get too wrapped around the axle about it. It is not Physics, Chemistry, or Math - it is cladistics. i.e., using data derived from techniques that originally gave us our cartoon view of earlier humans, and maybe later findings like DNA.

Hmm.

 

[pinball1970]

This question probably shows my ignorance about how genetics work.

And mine...
This question is more difficult than it looks and I studied this, to a point, going back a few years ago.

I will not expand on explanations from more qualified posters but I will post this link for you (further reading..)
A comparison between humans and Chimp genomes.
https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-020-06962-8/tables/1
The kind of things they compare with genomes.
You can click the numbers highlighted, they will provide a link to a detailed study/studies and PubMed will give you the abstract like this https://pubmed.ncbi.nlm.nih.gov/18562338/

We split from a common ancestor with Chimps about 7 millions years ago.