Proving the Neanderthal theory with 23andme

[rdos]

Yes, I think this is entirely possible, with quite a small sample. 23andme.com has a quite popular gene-test that in addition to the usual things also gives you a percentage of Neanderthal heritage. What is most interesting is that 23andme scans for almost 1,000,000 SNPs, and their Neanderthal heritage estimator uses them all. This is because they base it on factor analysis (PCA), which is the same method that Aspie Quiz uses to estimate neurodiversity scores.

Anyway, a month ago I inserted a question about 23andme score into Aspie Quiz, estimating a very low (but-non-zero) answer frequency. I was lucky as Aspie Quiz was featured on US radio on Friday a week ago, and this has generated a massive amount of answers (4,000 per day since). Because of this, I now have 71 answers from people of European descent that also did 23andme, and their Aspie score correlate 0.26 to percentage of Neanderthal heritage. One-sided p is about 0.015.

I've heard people complain in the past that "I cannot relate autism with Neanderthal", but now that I have almost done this, there are other objections. I hope that somebody will eventually do a controlled study on this. All that is needed is a population of perhaps 50 diagnosed individuals with ASD and a control group about the same size. OTOH, people will probably complain that a correlation of 0.26 is too small to prove the point. What the small correlation probably proves is that neurodiversity is not primarily to be found in SNPs, and that such research will never succeed.

[slartibartfastibast]

OTOH, people will probably complain that a correlation of 0.26 is too small to prove the point. What the small correlation probably proves is that neurodiversity is not primarily to be found in SNPs, and that such research will never succeed.

Even if autistic phenotypes are mostly due to CNVs or epigenetic stuff the technology to quantify it will be here within the decade.

There are alternate ways to link ancestry and neurodiversity:

Schizophrenia as variation in the sapiens-specific epigenetic instruction to the embryo (Jan 2012)

According to the hypothesis these sequence changes, together with one or more deletions and a paracentric inversion in the Y block, were successively selected; late events in this series established cerebral asymmetry (the ‘torque’) as the defining characteristic of the human brain.

As of January at least one published paper has referred to autism as a possibly atavistic (ancestral) variation:

Hominins and the emergence of the modern humanbrain (Jan 2012)

Autism, a neurodevelopmental disorder characterized by impaired social interaction for which a clinical definition is ongoing. is referenced in studies of fossil hominin brain structure and function, either as an analogy for developmental differences between closely related species or as a potentially atavistic indication of actual primitive phenotypes. For example, an autistic child lacking language created naturalistic artwork much like that from the Upper Paleolithic, on the basis of which it was suggested that fAMHS could have also lacked fully modern cognition (Humphrey. 1998).

If I'm reading that correctly, the theory that autism may be a selectively expressed ancestral phenotype is already on paper in a scientific journal. That lends the Neanderthal Theory considerable credibility.

My own hypothesis just extends the idea of marginally expressed ancestral phenotypes beyond autism. You also already did this, but I don't think you mentioned eusociality.

Here's my abstract:

ASDs (autism spectrum disorders) are hypothesized as one of many adaptive human cognitive variations that have been maintained in modern populations via multiple genetic and epigenetic mechanisms. Introgression from "archaic" hominids (adapted for less demanding social environments) is conjectured as the source of initial intraspecific heterogeneity because strict inclusive fitness does not adequately model the evolution of distinct, copy-number sensitive phenotypes within a freely reproducing population.

Evidence is given of divergent encephalization and brain organization in the Neanderthal (including a ~1520 cc cranial capacity, larger than that of modern humans) to explain the origin of the autism subgroup characterized by abnormal brain growth.

Autism and immune dysfunction are frequently comorbid. This supports an admixture model in light of the recent discovery that MHC alleles (genes linked to immune function, mate selection, neuronal "pruning," etc.) found in most modern human populations come from "archaic" hominids.

Mitochondrial dysfunction, differential fetal androgen exposure, lung abnormalities, and hypomethylation/CNV due to hybridization are also presented as evidence.

[rdos]

Even if autistic phenotypes are mostly due to CNVs or epigenetic stuff the technology to quantify it will be here within the decade.

Yes, but the main challenge lies in how to retrieve this from the Neanderthal genome, as they only have very short pieces of DNA today. They might be able to recreate small CNVs, but long will probably be out of reach.

As of January at least one published paper has referred to autism as a possibly atavistic (ancestral) variation:

Hominins and the emergence of the modern humanbrain (Jan 2012)

Autism, a neurodevelopmental disorder characterized by impaired social interaction for which a clinical definition is ongoing. is referenced in studies of fossil hominin brain structure and function, either as an analogy for developmental differences between closely related species or as a potentially atavistic indication of actual primitive phenotypes. For example, an autistic child lacking language created naturalistic artwork much like that from the Upper Paleolithic, on the basis of which it was suggested that fAMHS could have also lacked fully modern cognition (Humphrey. 1998).

Interesting.

If I'm reading that correctly, the theory that autism may be a selectively expressed ancestral phenotype is already on paper in a scientific journal. That lends the Neanderthal Theory considerable credibility.

I suppose we must first retire the old school before much progress will be made. I've been on a few professional mailing lists for several years now, and posting things related to the Neanderthal theory there. In the beginning, I was censored, but as they switched moderator I can post more or less freely. At this point I'm even getting occasional responses, so I know at least some people read what I post.

My own hypothesis just extends the idea of marginally expressed ancestral phenotypes beyond autism. You also already did this, but I don't think you mentioned eusociality.

Here's my abstract:

ASDs (autism spectrum disorders) are hypothesized as one of many adaptive human cognitive variations that have been maintained in modern populations via multiple genetic and epigenetic mechanisms. Introgression from "archaic" hominids (adapted for less demanding social environments) is conjectured as the source of initial intraspecific heterogeneity because strict inclusive fitness does not adequately model the evolution of distinct, copy-number sensitive phenotypes within a freely reproducing population.

Evidence is given of divergent encephalization and brain organization in the Neanderthal (including a ~1520 cc cranial capacity, larger than that of modern humans) to explain the origin of the autism subgroup characterized by abnormal brain growth.

Autism and immune dysfunction are frequently comorbid. This supports an admixture model in light of the recent discovery that MHC alleles (genes linked to immune function, mate selection, neuronal "pruning," etc.) found in most modern human populations come from "archaic" hominids.

Mitochondrial dysfunction, differential fetal androgen exposure, lung abnormalities, and hypomethylation/CNV due to hybridization are also presented as evidence.

I've naturally read your hypothesis, and found it very interesting. We are more or less on the same side, just with a little different interpretations, which I find quite healthy and useful. Now I have somebody that can challenge my own interpretation, which is nice.

[Sophist]

There is now the Human Epigenome Project, and it is becoming more and more feasible to map an individual's epigenetic info, so to combine that with the genome would be a great stride.

[rdos]

I made a quick check which questions correlated most with 23andme, and it strikes me that hunting / perception / talent on Aspie side, and compulsion / social traits on NT side are overrepresented. It seems like issues that can be related to modern humans and recent evolution in Neanderthal are most correlated to 23andme, which is kind of unexpected. These traits only explain a smaller percentage of neurodiversity (maybe 10-15%). The correlation with 23andme also seems to end up between 0.2-0.3. Perhaps the other traits are part of modern Africans because of early back-migration? I once made the hypothesis that modern humans formed after hybridization in Eurasia, which seems to be supported by this new finding.

[slartibartfastibast]

This question was deleted by yahoo right before I could craft an answer: http://webcache.googleusercontent.com/s ... 424AArGE8e
Screenshot: http://i.imgur.com/ZjnMF.png

Apparently it violated their community guidelines.

[rdos]

This question was deleted by yahoo right before I could craft an answer: http://webcache.googleusercontent.com/s ... 424AArGE8e
Screenshot: http://i.imgur.com/ZjnMF.png

Apparently it violated their community guidelines.

Yes, some people are scared about the truth. I just wonder why? Is it because they have commersial interests in cures, make the misguided mistake to think that behind every autistic is a dysfunctional NT or something else?

[mamasatya]

Looks to me like the poster was careless with his words: has any one actually took the time to look at a autistic and notice that they more resemble neanderthals? prominents brow ridges, wide philtrums, baby faced, with a somewhat blunted..

Maybe someone took knee-jerk offence and reported it and a guy in malasia on a dollar per day played it safe and hit delete.

There's no conspiracy to keep the neanderthal hypothesis down per se, it's just rather difficult to prove without being insanely offensive in some way. As demonstrated here: http://silencedbysilencedbyageofautism. ... /Aetiology

If you have a moment also, http://silencedbysilencedbyageofautism. ... teric.html

[rdos]

But the thing is that crazy hypothesis with no support at all can be listed at Wikipedia, get lots of media attention and eventually get published in peer-reviewed journals. Why are those not evaluated critically, and silenced if they cannot be proved?

Just take Wikipedia as an excellent example. The Asperger's syndrome article has been medicalized, and a few regulars will not except anything that is not published in the peer-reviewed literature. Any crazy hypothesis that is published can be added there even if it has been disproved, or has no support in research at all. Articles about neurodiversity has been repeatedly deleted (I'm not sure if there currently is such or not at Wikipedia). The Neanderthal theory was deleted several times, even when it was widely spread on the web, and actually refered to by some professionals. So was the article about Aspie Quiz.

And then most of the peer-reviewed journals have refused to send the Aspie Quiz article on peer-review, probably because editors are afraid of publishing it. This kind of goes in circles.

The thing is, even if I can prove that p < 0.01 that ASD and Neanderthal are related, it won't hit the media, it won't get listed on Wikipedia, and it probably cannot be published in the peer-reviewed literature because nobody wants to publish it. That's not how science should be done, but it seems to be how psychiatric "science" works.

[mamasatya]

No amount of gene correlating proves a causal link between Neanderthals and autism.

http://silencedbysilencedbyageofautism. ... 7758305002

All that you can show is that maybe autistic people and Neanderthals share those genes because of some environmental challenge in common (we hypothesis that the ability to prey on conspecifics when the local food-web goes belly-up is the conative drive of nature behind such evolutionary avenues of choice).

If you really want to prove that Neanderthals cause autism then you have to go to south Africa and start screening for populations where little to no 'autism' can be found, and then having found a suitable candidate population, get all of them, including any autism suspects, checked out both genetically and physically and behaviourally. One readily assumes that one can find suitably matched controls locally who have DNA from Neanderthals, to rule out 'living in the Congo' rather than 'No Neanderthal DNA' as the possible cause of 'No Autism'.

We hypothesise that autism is not found in populations who lack the usual ~ 2.5% Neanderthal DNA.

Whichever way we look at it, it's not an inexpensive study. I'd guess, millions of dollars, to answer your question.

[Sophist]

This is assuming that the roots of autism lie solely in genetics rather than epigenetic mechanisms.

Viewing genetics as a "cause" of anything reveals a poor understanding of biology and development in general.

[mamasatya]

This is assuming that the roots of autism lie solely in genetics rather than epigenetic mechanisms.

Viewing genetics as a "cause" of anything reveals a poor understanding of biology and development in general.

Quite right! The way I look at it, it's like saying the blueprint for a building is the cause of a building (or the reason for designing it a particular way). Even if you could pour building materials and fuel onto the blueprint and have the building magically spring into existence (DNA works this way, sort of) it's still not the 'cause' of the building.

Sophist how do you feel about the 'cannibal / hypernutrition / weather / no more cannibals' suite of hypotheses for autism?

[rdos]

No amount of gene correlating proves a causal link between Neanderthals and autism.

I disagree. Any statistically significant correlation between amount of Neanderthal DNA and diagnostic status of autism proves there is a link. There can be no other explanation for such a correlation except that genetic material from Neanderthal contribute to autism (which doesn't mean it solely causes autism).

All that you can show is that maybe autistic people and Neanderthals share those genes because of some environmental challenge in common (we hypothesis that the ability to prey on conspecifics when the local food-web goes belly-up is the conative drive of nature behind such evolutionary avenues of choice).

Impossible. If it can be shown that autistics, on average, have higher levels of Neanderthal inheritage, this can have no other cause than Neanderthal heritage contributing to autism. Autistics could not have accumulated exactly the same mutations as Neanderthals, especially not when we speak about a large amount of mutations, not just a few, if they didn't inherit them from Neanderthal.

If you really want to prove that Neanderthals cause autism then you have to go to south Africa and start screening for populations where little to no 'autism' can be found, and then having found a suitable candidate population, get all of them, including any autism suspects, checked out both genetically and physically and behaviourally. One readily assumes that one can find suitably matched controls locally who have DNA from Neanderthals, to rule out 'living in the Congo' rather than 'No Neanderthal DNA' as the possible cause of 'No Autism'.

That's just as impossible. South Africa has 10% European ancestry, which means we expect to see Neanderthal DNA and neurodiversity in South Africa.

We hypothesise that autism is not found in populations who lack the usual ~ 2.5% Neanderthal DNA.

Not only do we hypothesize that Neanderthal DNA is present in Africa, but the latest check on which traits are most correlated between the Neanderthal DNA test and neurodiversity, indicates that the Neanderthal DNA test severely underestimate the amount of Neanderthal / Eurasian archaic DNA in Africa.

[rdos]

This is assuming that the roots of autism lie solely in genetics rather than epigenetic mechanisms.

Not so. Currently, the 23andme test can only explain 10% of neurodiversity. We still lack 90%. OTOH, the present result is not evidence that 90% of neurodiversity is environmental or epigenetic. Other possible reasons for the low amount is that 23andme only uses SNPs, and not CNVs. A test that compared CNVs could potentially get a much higher correlation that might explain much of neurodiversity. About the only thing we know is that at a minimum Neanderthal DNA can explain 10% of neurodiversity. Then we have the difference between ASD diagnosis, and neurodiversity as defined by Aspie Quiz. They have a little above 0.8 in correlation, but that means that they don't share 30-35% of the traits. These 30-35% are probably mostly environmental issues.

[mamasatya]

Rdos maybe I can illuminate things a little more:

All higher life forms have an option that we are not particularly familiar with: Cannibalism.

In a situation where conspecific predation becomes necessary for survival, we start to realise that there are really two options for conspecific predation: Post-mortem or Pre-mortem.

If these two options are available to 100,000 different creatures, then there can be many more than 100,000 ways to express the difference between these two options genetically.

We hypothesise that if neanderthals were predisposed to pre-mortem conspecific predation and paraspecific predation of other hominids such as homo erectus, then they would exhibit genetic markers corresponding to the hominid conspecific predation specialist (as opposed to the genes for dog-eating-dogs), and we further hypothesise that the 'triad of impairments' found in autism is one such aspect of such predisposition and corresponding genetic markers, in common with those found in neanderthals, may be found to be active within autistic populations.

I can catch a whole herd of woolly mammoth with just a jam-jar and a sock, and those thick brow-ridges are naturally selected for not loosing one's vision when getting in a fist-fight with a cave bear.

How well can you *really* know your inner neanderthal if you've never been in a fight with a cave bear and how can you be sure there aren't a few neanderthals still left alive, in Tibet with internet connections, rather offended by your theory that they are autistic?