I’m absolutely certain it’s incorrect. The average of shared genes between an ancestor and descendant, ignoring those shared with lots of ancestors, drops to something insignificant in a few generations, and that’s just the averages. If you could go back just ten generations and check the 1024 ancestors you had back then (ignoring incidences of pedigree collapse) you wouldn’t find a nice even distribution of genetic relatedness and it’s unlikely a distinct subgroup among them would contribute to your genetic makeup in proportion to their ratio in that generation.
These are mostly a bunch of silly “just so” stories. Epicanthic folds do not in any way shield the eyes from sunlight. If anything, Asian people tend to have eyes that are less deeply recessed than Caucasians and therefore more exposed to direct sunlight. And snow dwellers would mostly need to be protected from the sunlight reflecting up the ground, so they’d need reverse epicanthic folds.
I’m almost surprised you didn’t also surmise that people with Down’s syndrome evolved epicanthic folds to protect them from accidentally poking their eyes out with a fork.
I disagree. Yes, if you go back 1500 generations (in the ballpark for Neanderthal interbreeding), each ancestor slot will contribute only 0.0000000000000000…0000001 base pairs on average, and I didn’t type enough zeros in that fraction. Very close to zero, but not exactly zero as a mean expectation. The law of large numbers comes to the rescue! If 4% of your ancestor slots were Neanderthal, 4% of your DNA bases will be Neanderthal even though each slot contributes “about” zero!
I think I was correct that “4% Neanderthal DNA” means that 4% of your ancestor slots 50,000 years ago or so were Neanderthal. I’ve no idea how they make that calculation.
Why do you think you’re correct? Can you link to a cite that validates that interpretation? Not saying I know you’re wrong, I’m just wondering why you are so certain you are right.
Here’s a paper describing 23andMe’s methodology for estimating Neanderthal ancestry:
https://23andme.https.internapcdn.net/res/pdf/hXitekfSJe1lcIy7-Q72XA_23-05_Neanderthal_Ancestry.pdf
Mainly just common sense and interpretation of English. What else could it mean? (Does 4% “seem” high? Remember that the inheritance pathway was riddled with bottlenecks and “founder effects.”)
(BTW, I didn’t fathom your earlier response to me in this thread. You seemed to correct me on the distinction between genes and base pairs even though my post mentioned neither.)
A plain English interpretation of a complex scientific matter that was reported in the popular press is not necessarily the right answer. In fact, I would caution against it.
I wasn’t corrected you-- I was elaborating on what I posted. Most people think in terms of genes, but my understanding of the comparison is that the comparisons is between base pairs.
It’s a rough estimate of how many base pairs in the whole DNA chain were contributed from genetic material introgressed into (mostly) out-of-africa lineages.
In theory, it would roughly correlate with “genes” but the word “gene” is used very loosely, and there isn’t an enforced standard way to use it. And it’s not as if we can take the whole DNA chain and say, “From here to here is Gene X that does Function Y” all the way along the chain.
The 1-4% number you see for eurasian populations (and probably highest in asians) is an amount of genetic material that would not be there if out of africa human lines had not reproduced with Neandertals. (and for asians, maybe, more than one pulse of hanky panky 
)
DNA is a big molecule, and what we call “genes” are really location references where we think a given range of base pairs drives a particular physiologic process (by coding for production of proteins blah blah blah) . As humans, we all have about the same basic (physiologic) processes, so it becomes a very tricky distinction to say a given gene is “different” (and, more precisely, a given gene variant).
Sometimes a single base substitution drives a known, measurable difference. Sometimes nothing happens. Paul Thompson looked at a cohort of 20,000 people and found that a single cytosine for thymine substitution in a single gene might drive as much as 1% difference in average IQ. But most of the time it’s impossible (without huge cohorts looking at very narrow changes) to figure out what the hell the gene variants are even doing, much less figure out if they are “different” in a functional sense. We do it for diseases, because it is important there, and for a handful of other areas where we are just curious.
A lot of DNA is just junk; left over crap. Some gene variants get turned on and off. Most of 'em probably affect each other in totally unknown ways, or have effects way beyond the effects we are sure we’ve worked out.
The fact that such a high percentage (1-4 % of genetic material is very high when you consider that a single base pair change might be significant) of Neandertal genetic material survives is interesting, but by itself does not tell us if it makes one set of human lines actually different from another set in a qualitative way. It does mean that out of africa human lines have DNA from both the “original” anatomically modern human ancestor a couple hundred thousand years ago in africa, plus DNA from even more archaic lines coming in through Neanderthals, Denisovans, and the like.
Still, one good gene variant is worth a thousand lame ones, so I’m not sure this idea of percentage difference means much as a way of quantifying differences among human lines.
What you really want to look for are gene variant prevalences that can be shown to be driven by natural selection. IOW, genetic variation that matters to Nature in driving reproduction. My 2c, anyway.
By CP:
"The 1-4% number you see for eurasian populations (and probably highest in asians) is an amount of genetic material that would not be there… "
Ok, correcting my own crap here. This is a terrible way to put the concept I’m trying to get across b/c it implies a pure addition.
Is this any better?
…an amount of genetic material that is unique to lines which are descended from recent ancestral pools that mated with Neandertals…
The sense in which it is “unique” is beyond the scope of a post here, but think of two groups of people, one of which can call Neandertals “great-gramma/grampa” from 50 kya, and the other of which has to go back to a much more ancient ancestor common to both Neandertals and anatomically modern humans (L0s, for mtDNA markers) to get to the same set of great-grams and gramps.
No. It’s not as if a geneological tree keeps expanding. You’d have more “ancestor slots” (?) than the amount of people who ever existed.
Tiny little founder groups might have one or two boinkings (more for the asians, maybe; see my reference above), or might have just been having fun all the time.
What we think right now is that however the introgression happened mechanically, the residual average total base pair contribution can be estimated at 1-4%. But of course those base pairs don’t get transmitted individually; they come in as genes and other large preserved chunks–some chunks that might be really good gene variants; some that might just be hitching a ride.
You couldn’t look at one base pair and say, “that’s Neandertal!”
To say 4% of your “ancestor slots” were held by Neandertals is not a good way of putting it.
Despite your “No”, we might be in agreement. “Ancestor slot” is precisely the correct way to put it. Yes, there are well over a googol slots at the time depth we speak of --* So what?*
And, yes, the “founder effects” mean that it’s hard to deduce the detailed scenario of interbreeding from the genetic evidence. But I stand by my claim that “4% Neanderthal” means that, at an appropriate time-depth, 4% of the “ancestral slots” were Neanderthal. What else could it mean?
The mean expectation is correct, but the more generations, the more spread out that distribution will be. I’m not very good at stats, so I did a simple simulation to check my intuition.
Imagine you have a couple made up of one Neanderthal and one Sapiens, all their children, generation 1, will be 50% Neanderthal.
My simulation then uses a binomial distribution with 100 genetic variations, each with a 50% chance of being inherited and looks at 1000 generation 2 children.
The average genetic Neanderthalness of those is as expected 0.25, but the standard deviation is 0.025, that’s 0.25±10%
30% of that generation will have less than 23% or more than 27% Neanderthal genes.
In the third generation it’s ±14%
In the third generation it’s ±18%
In the third generation it’s ±21%
In the third generation it’s ±23%
This assumes no other Neanderthal contributors along the way, while in reality even with just one Neanderthal contributor you’d start seeing interbreeding after a handful of generations, but you’d be just as likely to interbreed with a distant cousin with a Neanderthalness on the low end as a really Neanderthaly one, and that should contribute to the smear in the same way that you’re as likely to get more Neanderthal genes from your parents as you are to get less.
Now if by “* “4% Neanderthal DNA” means that 4% of your ancestor slots 50,000 years ago or so were Neanderthal.*” you have meant .04±99% all along, I’m sorry for the misunderstanding. And if I’m wrong in my intuition, I’m ready to be educated.
At 3.2E9 base pairs, assuming recombination anywhere, and completely homogeneous passing of genetic material (both assumptions are bogus but let’s pretend), 31 generations back would get you one base pair from each parent[sup]n[/sup]. Finding 4% Neandertal base pairs would be consistent with ~4% of those n = 31 parent slots being 100% Neandertal.
But of course, recombination is chunky. The probability of having zero DNA from a given ancestor goes to unity pretty darn quick (~12 generations?). Some genes are more likely to get passed on.
Well, it can mean “ancestral slots” if you make a hypothetical model construct that has no reality and use “ancestral slots” any way you want.
What it means in real life is that 1-4% of the genome of modern eurasians reflects a pattern that is derived from the Neanderthal genome and not from the “anatomically modern” source pool genome that arose in africa about 200 kya (the L0 mitochondrial line).
When we look at a given base pair in our genome, the only pattern we see is that cytosine and guanine pair up, as do adenine and guanine. If we step back, though, we see larger patterns which are created because whole chunks tend to move as a unit as they are passed from ancestral to descendant organisms. Entire sequences of AG/CT pairings along the molecule are preserved, in part because it requires a whole nucleobase range to execute a coding function.
We call those ranges a “gene” when we can identify a reasonable nucleobase start and end point, and have an idea what function is created. There are other chunks (DNA ranges) which move together as a unit but are either junk (we think there is a lot of that) or unidentified genes, or whatever.
Nevertheless, because sections of DNA are preserved as a unit (whether “genes” or not), patterns are preserved. What we call a gene has multiple variants as individual nucleobases within that gene sequence location are swapped out by random changes, some of which are preserved because evolution finds them suitable to drive reproduction. Over time, the patterns (DNA sequences) change among separated groups.
More precisely, the gene variants and other sets of DNA sequences reflect nuclebase changes derived and preserved for one reason or another along the descendant lineages. What happens over time is that the prevalence of newly introduced patterns (sequence variations) changes from group to group.
We use various marker techniques to track these patterns, and infer from those techniques which patterns derive from which ancestral lineages.
We can therefore also infer from pattern sampling (and, sometimes, more direct measurements) what total percentage of DNA chunks are derived from which ancestral group.
For Neanderthal contribution, about 1-4% of the genome of modern eurasions show DNA sequence patterns that existed only in Neanderthals.
That’s what it means. Nothing about “ancestral slots,” and I have not seen papers which discuss it that way.
What will happen as further work is done is that we will figure out which areas of physiology are involved, and perhaps eventually what the exact effect was.
Nice diagram here.
“The team identified more than 70 gene changes that were unique to modern humans. These genes are implicated in physiology, the development of the brain, skin and bone…
Any functional significance of these shared genes remains to be determined, but that will certainly be a focus for the next stages of this fascinating research.”
adenine and thymine… AT/CG
sheesh.
Epicanthic folds are useful in any condition which is improved by squinting. Harsh wind, extreme low temps, extreme high temps, extremely dry air, bright light, all these are made easier to deal with if you can provide more protection to the eyeball with a minimum of effort and vision loss.
There is a great deal of argument as to whether it originally emerged as an adaptation to cold, or to bright sunlight. I was unable to locate any theories in relation to flatware.
As for Down’s syndrome, it’s an anomaly and not adaptive in any way that I have ever heard. Although, come to think of it, it does tend to teach siblings compassion and cooperation. At any rate, I’ve known quite a few people with Down’s, and none of them had any difficulty with eating utensils.