DNA ethnicity tests

[SamTheSkeptic]

First off hello everyone, I am glad to have joined this forum and look forward to being a contributing member here.

 

My first post is regarding the "DNA ethnicity tests" that many have probably heard of already. These tests analyze your DNA and as

far as I have seen, provide results which are traceable ultimately back to 4 major categories: Sub Saharan African, European, Asian, and

Indigenous/Native American. What got me thinking was a show that I had seen a while back on PBS, which featured a bunch of well known

stars who had these DNA tests ran on them. Many were surprised by the results, which often included categories that they didn't suspect

(I specifically remember A Mexican actress who was rather shocked to learn that she had deep Asian roots). What I noticed was, that some

people who were tested had 0% Sub Saharan African in their ethnic makeup. How could this be if we all originated in Africa? I don't doubt that

the human race did indeed originate in Africa, but it does seem a bit strange that some people exhibit no traceable lineage back to it; shouldn't we

all theoretically have a little bit of "African DNA" in us? How is this possible? Is it just that the DNA is so diluted by now that it simply doesn't

show up anymore using the current test methods?

 

Any thoughts?

 

Cheers.

[Mr Skeptic]

Think of it like this. Suppose you have a species of flower that can have red, white, blue, and purple flowers. Then someone takes the wild population and divides it into groups based on the color. Later on, you could do a genetic analysis of the DNA and determine which group an individual plant belongs to (or you could just look at the color). However, that grouping was rather arbitrary, and you could as easily grouped them by height, pest resistance, seed size, or whatever other attributes you can think of. Racial genetics is rather like this, but not so much based on specific genes that either exist or not in a population, but rather the frequency of an enormous range of genes. One assortment of gene frequencies we call "african", another "european", "asian", or "native american". You could take any of these populations and through careful breeding change the gene frequencies to match those of another race, unless there are some genes unique to any one group, but that would take a very long time. Because so many genes are involved, for quite a while it would be possible to identify the proportions of the groups which form your ancestry.

 

Another example would be if you had four children. One of your children stayed at home and the rest went each to a different continent, and all had children of their own. By DNA testing you can tell which of your children the grandchildren belong to, even though they are all descended from you. You could also tell that one of the grandchildren was not the son of your stay-at-home son, which of course should be no surprise.

 

So humanity divided from its common origins into a few groups each of which went their separate way genetically with low intermingling. It should no more surprise you that the tests might show no african ancestry in an euoropean than if it shows no euoropean ancestry in an african, since all are from a common origin.

[SamTheSkeptic]

Think of it like this. Suppose you have a species of flower that can have red, white, blue, and purple flowers. Then someone takes the wild population and divides it into groups based on the color. Later on, you could do a genetic analysis of the DNA and determine which group an individual plant belongs to (or you could just look at the color). However, that grouping was rather arbitrary, and you could as easily grouped them by height, pest resistance, seed size, or whatever other attributes you can think of. Racial genetics is rather like this, but not so much based on specific genes that either exist or not in a population, but rather the frequency of an enormous range of genes. One assortment of gene frequencies we call "african", another "european", "asian", or "native american". You could take any of these populations and through careful breeding change the gene frequencies to match those of another race, unless there are some genes unique to any one group, but that would take a very long time. Because so many genes are involved, for quite a while it would be possible to identify the proportions of the groups which form your ancestry.

 

Another example would be if you had four children. One of your children stayed at home and the rest went each to a different continent, and all had children of their own. By DNA testing you can tell which of your children the grandchildren belong to, even though they are all descended from you. You could also tell that one of the grandchildren was not the son of your stay-at-home son, which of course should be no surprise.

 

So humanity divided from its common origins into a few groups each of which went their separate way genetically with low intermingling. It should no more surprise you that the tests might show no african ancestry in an euoropean than if it shows no euoropean ancestry in an african, since all are from a common origin.

 

Mr. Skeptic,

 

Thank you very much for the reply, however I don't think that you really addressed the question... I am aware that certain gene frequencies are what comprise the "categories" if you will (like Sub-Saharan African or European), but the question is: "WHY is the Sub Saharan African gene frequency completely absent from the DNA tests of certain individuals, if all humans descended from ancestors in Sub Saharan Africa?"

 

Your last statement particularly troubled me. . .

It should no more surprise you that the tests might show no african ancestry in an euoropean than if it shows no euoropean ancestry in an african, since all are from a common origin.

 

African ancestry should show up in all humans theoretically, but the reverse is NOT true. Why would you expect to find European ancestry in an African person's DNA if they come from a family line that never intermingled with (the more newly evolved) European people? The European DNA is "newer" in the sense that it evolved later in time, and so theoretically you wouldn't expect to find traces of it in the "older" Sub Saharan African DNA.

[Mr Skeptic]

No, species do not get stuck in time nor do species stop evolving. Why do you expect modern Europeans to have something in common with modern Africans, that the modern Africans do not have in common with the modern Europeans? Yes we all share ancestral DNA but that is not what this test is measuring. We have DNA in common with all species all the way down to bacteria. And we have genes separate from them too. But these tests are of modern Africans and modern Europeans, not modern Europeans and time-travel-back-a-million-years Africans. And even if you did test against time-travel-back-a-million-years Africans, the result would be the same -- the genes common to both get ignored for the test, and the different genes get used to compare which group someone belongs to.

[SamTheSkeptic]

There's no need to travel back a million years, or even a few hundred thousand. There are many "modern" African people who show nearly 100% Sub Saharan DNA on their test results. . .this is what i refer to as "old" DNA and indeed why I put it in quotations. Obviously, it isn't frozen in time, but these people are closer genetically to the first humans on the evolutionary tree than any European/Asian/Native American person. . and therefore it would be surprising to find "European DNA" in their test results. However, it should then follow that this Sub Saharan African DNA SHOULD be found in ALL other humans. . unless of course the tests are not accurate enough to encompass such a long time period and the genetic frequencies have been greatly diluted (which I believe to be the case).

[NTettamanti]

I think your last sentence answered your question; however, it may be possible that the tests only report the unique traits of each group. I think Mr. Skeptic points it out best when he claims "the genes common to both get ignored for the test, and the different genes get used to compare which group someone belongs to."

 

Of course there is going to be shared DNA between all humans; if they were testing our entire genome then the results would show that we are all incredibly similar. That's not what they were looking at though, I would imagine that they would be taking genes that exist in large frequencies from every geographical area and compare those genes against that of the stars.

[Mr Skeptic]

Very well then, suppose you know of DNA that all humans hold in common. Could this DNA be used to distinguish between groups of humans? If so, how? If not, why not?

[lemur]

Surely everyone would share some genes with the sub-Saharan African gene index if the statistical parameters were relaxed enough. I can't imagine that there is some gene sequence that absolutely only occurs in people whose ancestors ever lived in that region and others that only ever occur in people whose ancestors never lived in that region. If so, this would be a basis for re-establishing the idea of race as a biological form of sub-species; but if that were the case, how would all humans be able to reproduce regardless of racial identity? And, since they can all reproduce with each other, what is the point of creating distinctly classified genetic origins? Are people getting back into the market for racial exclusivity? Are governments going to start genotyping individuals into races and restricting "interbreeding?"

 

Regardless, the interesting thing about these genetic origins, imo, is what basis there is to assume that these regional classifications represent some stable population regionalism? Didn't humans basically spend pre-history wandering around anywhere and everywhere they could go? If so, what would account for these regional gene indexes? How much regional population stasis was actually occurring and why? Did people tend to be nomadic in recirculating patterns and reproduce only in certain places with certain people? Didn't men always wander around and "sow their seeds" wherever they could?

 

Is it really fun to study race in this way? To me it just seems like an extension of 19th century raciology. It brings up mostly sad thoughts of the way people used to treat others (and still sometimes do) on the basis of racial difference.

Edited February 16, 2011 by lemur

[SamTheSkeptic]

Okay, let me back up a bit here... and thanks for chiming in too Ntettamanti.

 

I of course understand that all living things have DNA. And I also understand perfectly that these "ethnicity" tests are ultimately

just tests which look for gene frequencies that express different variations among members of our species, homo sapien sapien.

And yes, I also understand that the tests are looking for differences rather than similarities, otherwise the whole purpose of the

tests would be defeated.

 

Now with that being said, let me illustrate my point through an example:

 

The first humans came onto the scene in Sub Saharan Africa, then large groups of this original blood line migrated to Asia and Europe.

These groups settled in and evolved over time, procreating and passing on new mutations/dominant genes which resulted in a

group of people who eventually looked significantly different from their original group in Africa. Now of course the "original" group

in Africa, who never migrated, didn't stop evolving or stay stuck in time genetically. However, they still retain much more of the original blood line

(referred to as "Sub Saharan DNA") than the groups who migrated off. . .

 

At this point, how could you expect one of the direct descendants of the original group to have any of the more newly evolved genetic

frequencies that show up in the groups who had migrated? Another example is with humans and chimps:

We share 98% of our DNA with chimps. . .that other 2% is what makes a very large difference in terms of our physical features and brain

capacity. Now, we know that chimps are "further back" on the evolutionary tree than humans. . they are closer to the common ancestor than we are; or in other

words, our species is newer because it came into existence more recently.

 

So it follows then that you would expect to find "chimp DNA" in human tests. . .but you would NEVER find that other 2% ("Human DNA") when testing a Chimp!

Time only moves forward, never backwards. . .

[NTettamanti]

The direct descendants may not have the same DNA as the groups who have migrated; however, I don't believe that just because there are changes in the DNA sequences that it necessarily means that those newly evolved frequencies are any better or worse than the original. Also, this would simply show that there would be genetic variation from one group to another (that can be quantitatively and qualitatively measured). We could expect that the new

 

Also, if chimps and humans have 98% DNA then you wouldn't find Chimp DNA in humans. What you would find is the DNA from our common ancestor; the 2% difference works both ways. To demonstrate this better, your statement could be rewritten as: you would never expect to find human DNA in chimp tests... but you would never find that other 2% (chimp DNA) when testing a chimp. That's the point; we are 98% similar and the 2% differences work both ways.

 

I think Lemur brings up really good points; particularly about the actual application of this science into society. I do believe, however, that these types of tests can actually destroy that harmful racial ideology; if people understood how similar they were then there would be no reason to engage in racial discrimination. In addition, even if individuals wandered everywhere, they eventually settled down in an area and the mechanism of migration took affect. I believe its clear that certain regional groups probably contain a higher frequency of certain genes simply due to a lack of migration. And while the idea may not be perfect, I think that the regional classifications we create can be (not necessarily are) descriptive of the gene pool in a particular area; however, I would say that these classifications are not prescriptive (they may change in the future) and they don't necessarily describe any important difference.

Edited February 17, 2011 by NTettamanti

[Mr Skeptic]

Another example is with humans and chimps:

We share 98% of our DNA with chimps. . .that other 2% is what makes a very large difference in terms of our physical features and brain

capacity. Now, we know that chimps are "further back" on the evolutionary tree than humans. . they are closer to the common ancestor than we are; or in other

words, our species is newer because it came into existence more recently.

 

Nope! The whole concept of one contemporary species being "newer" or "superior" to another goes against everything evolution says, and is reminiscent of the ancient Greek and Lamarkian ideas of a Great Chain of Being. One species may be closer to its ancestral species but that does not make it any worse nor better. You cannot ever tell by comparing two individuals which is closer to the ancestral species. To be able to claim similarity to a common ancestor would require either comparison with several other current related species or with fossils from the ancestral ones.

 

So it follows then that you would expect to find "chimp DNA" in human tests. . .but you would NEVER find that other 2% ("Human DNA") when testing a Chimp!

Time only moves forward, never backwards. . .

 

No... You will no more find "chimp DNA" in humans than you will find "human DNA" in chimps. You will find "common ancestor DNA" in both, in addition to finding "chimp DNA" in chimps and "human DNA" in humans. Even if you show that the chimps are closer to the common ancestor than are the humans, the shared DNA will still be "common ancestor DNA". All it would mean is that the species closer to the common ancestor would have a little more "common ancestor DNA" and a little less "new species DNA" than would the species less similar to the common ancestor.

[NTettamanti]

Yup; we're making the same point... I was just illustrating it in a poor way I was saying that the 2% difference exists because we have drifted away from our common ancestor, and saying that the idea that you won't find chimp DNA in humans is relative; because it works both ways.

 

To illustrate (perhaps in a better way), the x would represent DNA that was from the common ancestor, while H and C would be human and chimp DNA respectfully. You wouldn't ever find human DNA in chimps, and vice versa (because that is what makes us different!)

[xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxH]

[xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxC]

[SamTheSkeptic]

Well said NTett, I fully agree that the more newly evolved genetic frequencies aren't any "better" or "worse" than the older ones...this isn't the question at all here. Nor am I on any sort of racist agenda trying to prove genetic purity or dominance.

 

With regards to the Chimp example: of course this is a bit extrapolated, since I am comparing different species (chimps and humans) to varied members of the same species (African Vs. European). And you are certainly right in your points, I should have worded it better. The 98% similarity in DNA between humans and chimps is what we have from our common ancestor; in other words, we didn't evolve directly from chimps. But even then, the point here is that we are expecting to see this 98% similarity when we test any two chimp and human subjects. . just as we would expect to see some traces of Sub Saharan African DNA leftover in any two different humans tested. . .

 

Nope! The whole concept of one contemporary species being "newer" or "superior" to another goes against everything evolution says, and is reminiscent of the ancient Greek and Lamarkian ideas of a Great Chain of Being. One species may be closer to its ancestral species but that does not make it any worse nor better. You cannot ever tell by comparing two individuals which is closer to the ancestral species. To be able to claim similarity to a common ancestor would require either comparison with several other current related species or with fossils from the ancestral ones.

 

The "concept" of a species being newer (I never used the words 'better' or 'superior') is not a concept at all, it is a fact. Sharks for example, are much much older than humans. Our species' genetic origins only have their roots dating back to about 500,000 years ago. . .this is most definitely "newer" when compared to any given species of shark for example. You are right however, that you can't ever be 100% sure that any given species is genetically closer to its ancestral species. . but in some cases there is considerable archaeological/fossil evidence to support it.

Edited February 17, 2011 by SamTheSkeptic

[NTettamanti]

But even then, the point here is that we are expecting to see this 98% similarity when we test any two chimp and human subjects. . just as we would expect to see some traces of Sub Saharan African DNA leftover in any two different humans tested. . .

 

I assume we would as well... I guess it depends on the methodology of how they determined what genes were present in what area. The genes that all human share due sharing a common ancestor in our lineage are probably thrown out of the test results. I would assume, though I have no idea, that they simply take the genes that appear in high frequencies from different regions and compare them to other groups. If gene A exists in high frequencies in region 1 and 2 but not 3, then we can (probably not very accurately) assume that an individual who lives in region 2 probably originated from region 1 (assuming that most migrated from region 1, sub-Saharan Africa). Like I said, I have no idea how they actually calculate those numbers though.

[Mr Skeptic]

Well said NTett, I fully agree that the more newly evolved genetic frequencies aren't any "better" or "worse" than the older ones...this isn't the question at all here. Nor am I on any sort of racist agenda trying to prove genetic purity or dominance.

 

With regards to the Chimp example: of course this is a bit extrapolated, since I am comparing different species (chimps and humans) to varied members of the same species (African Vs. European). And you are certainly right in your points, I should have worded it better. The 98% similarity in DNA between humans and chimps is what we have from our common ancestor; in other words, we didn't evolve directly from chimps. But even then, the point here is that we are expecting to see this 98% similarity when we test any two chimp and human subjects. . just as we would expect to see some traces of Sub Saharan African DNA leftover in any two different humans tested. . .

 

And we do. We just ignore it for the test in question. Looking at similarities doesn't help you tell the difference.

 

The "concept" of a species being newer (I never used the words 'better' or 'superior') is not a concept at all, it is a fact. Sharks for example, are much much older than humans. Our species' genetic origins only have their roots dating back to about 500,000 years ago. . .this is most definitely "newer" when compared to any given species of shark for example. You are right however, that you can't ever be 100% sure that any given species is genetically closer to its ancestral species. . but in some cases there is considerable archaeological/fossil evidence to support it.

 

How are the Selachimorph any older than us Teleostomi? We were both Gnathostomata before we went our separate ways (and we still are). Of course species Homo sapiens is more recent than superorder Selachimorph. But I bet there's newer shark species than Homo sapiens. And certainly any shark species of your choice will be younger than us Teleostomi.

[SamTheSkeptic]

And we do. We just ignore it for the test in question. Looking at similarities doesn't help you tell the difference.

 

No, we don't. . .which is why I originally made the post. We don't always find the Sub Saharan African DNA in individuals, when theoretically we should.

Some people show up 0% Sub Saharan African. Why this occurs has yet to be explained.

 

 

 

 

How are the Selachimorph any older than us Teleostomi? We were both Gnathostomata before we went our separate ways (and we still are). Of course species Homo sapiens is more recent than superorder Selachimorph. But I bet there's newer shark species than Homo sapiens. And certainly any shark species of your choice will be younger than us Teleostomi.

 

I'm certainly not an expert on sharks, but I don't have to be to know that some species of shark evolved hundreds of millions of years ago...and that our species isn't anywhere near that old. Therefore, we are newer species in comparison.

[Mr Skeptic]

No, we don't. . .which is why I originally made the post. We don't always find the Sub Saharan African DNA in individuals, when theoretically we should.

Some people show up 0% Sub Saharan African. Why this occurs has yet to be explained.

 

It's been explained to you plenty. It's because we ignore 99.9% of our DNA when doing that test, and focus on that 0.1% that is different.

[SamTheSkeptic]

It's been explained to you plenty. It's because we ignore 99.9% of our DNA when doing that test, and focus on that 0.1% that is different.

 

It hasn't been explained at all. Think about what you're saying. How can Sub Saharan African DNA be "ignored" if it is one of the categories ('differences' as you put it) which are being tested for? If it was being ignored, it wouldn't even be on the test! I think what you're failing to see is that it is within this 0.1% that the differences exist...

Edited February 20, 2011 by SamTheSkeptic

[lemur]

I think Lemur brings up really good points; particularly about the actual application of this science into society. I do believe, however, that these types of tests can actually destroy that harmful racial ideology; if people understood how similar they were then there would be no reason to engage in racial discrimination. In addition, even if individuals wandered everywhere, they eventually settled down in an area and the mechanism of migration took affect. I believe its clear that certain regional groups probably contain a higher frequency of certain genes simply due to a lack of migration. And while the idea may not be perfect, I think that the regional classifications we create can be (not necessarily are) descriptive of the gene pool in a particular area; however, I would say that these classifications are not prescriptive (they may change in the future) and they don't necessarily describe any important difference.

I see your point about DNA showing that genetic/biological difference is negligible, but the problem is that people aren't generating these classifications for comparison out of thin air. They are the latest iteration of raciological classification, which tends to work because it has been a influencing factor in reproductive partner choice for some time. When the "2% genetic divergence" between humans and chimps is being discussed, do you notice that there is little attention paid to the fact that this difference must have been 1% at some point and 0% at some point before that when the two had a common ancestor? Then the question becomes what led primordial chimp-humans to begin selective breeding in a way that caused them to differentiate as breeding populations? Obviously, all species differentiation is evidence of what could be called "racism" in nature, but the question is whether this validates racism as a basis for mate-selection. I don't think it does any more than I think pure-breeding dogs is somehow contributing to the lives of individual dogs and the richness of their gene pool. So the question is what is the purpose of analyzing human genetic code for evidence of prior segregation would be except to re-institute such segregation in the form of supposedly "natural" population pluralism?

 

Still, in a cynical way I guess I would be interested to know how rigorously partner-selection would have to be constrained and for what period of time to generate the amounts of genetic differentiation present in contemporary humans. Is this really a question of genetic differentiation, though, or are genetic archaeologists just finding specific "index" genes that can be used to statistically correlate among people who live in geographical proximity? In that case, the point would be so much to identify significant genetic differences as it would be to seek genetic evidence of relative regional reproductive autonomy and trace this to identify individuals globally in terms of relatedness to regional populations, right?

 

So, if you wanted to, you could also perform statistical correlations by identifying global categories such as "people who can sing in key" or "people who like to dance" and identity genetic markers that correlate with the identified trait and then invite people to test themselves for genetic "linkages" to such categories? In that case, isn't it the a priori categories choses by the researcher rather than the genes themselves that decide which genes are compared statistically?