Question in regards to admixtures.

Irakli

Junior Member
Messages
8
Reaction score
0
Points
0
Ethnic group
Georgian
Hello, I am somewhat new to this subject although I have been reading for hours about the matter for the last two days and I was aware of haplogroups for many months. When I "set out" to better understand the genetic make-up of Georgians I was very often encountering a lot of discussions about the proportions of admixtures in certain ethnicities, like 20% ANE and 80% EEF in Georgians or theories like Georgians being mostly Caucasian Hunter Gatherers and EEF and an unending list of theories basically. But this question is not about Georgians specifically but about the meaning of admixtures. What are admixtures precisely? When I first got into this subject I had the impression that Y-DNA and mtDNA are the sole or at least primary defining factors when it comes to national genetics and phenotype, but as I am aware now autosomal DNA is a more important factor and it combines the genetics of all our ancestors instead of father's father's father's genes and mother's mother's mother's genes.

So are admixtures a group of certain subclades? A certain proportion of Y-DNA haplogroups+mtDNA haplogroups? I have seen some admixture maps on this website and a lot of them added up to more than 100%, so are admixtures pretty much arbitrary? You take one ancient genetic sample and compare them to other groups? Like the West Asian or Caucasian admixture for example, it said that Georgians had around 90% of it, and every other region had less than that. Does that mean that other regions' populations are discussed from the point of view of being similar to Georgians or is the Caucasian admixture a very real and objective fact? Because on the eurogenes website I have read through so many discussions and opinions that I feel mind****ed at this stage and the topic of population DNA seems to be pointless speculation. Also the mtDNA in pretty much every ethnic group is so diverse that I struggle to understand how you could attribute a certain haplogroup to a certain ethnicity or a group of ethnicities. Do admixtures have a better correlation with phenotypes or with actual genetic make-up of a group of people?
 
I understand that when two human groups live isolated for a long time in which they become genetically homogeneous, and then mingle with each other, a mixture of genes appears. I think, this mixture refers to autosomal DNA. I'm not clear either if for example when you say the "Caucasian admixture", it is with reference to the genetically homogeneous group from Caucasus or its combination with another different group? But I think it's most likely about the homogeneous genetic group from the Caucasus.
 
Uniparental markers like yDna do not have "necessarily" anything to do with autosomal signatures, although often they do. You have to look at each case individually.

Take, for example, Cameroon herders of Africa, who carry R1b but are autosomally West Africans. When a group of predominantly male migrants enters a region and mates with local females, their autosomal signature can disappear very quickly in their descendants.

Cameroon_2.jpg
 
(...) When a group of predominantly male migrants enters a region and mates with local females, their autosomal signature can disappear very quickly in their descendants.
Y chromosome is transmitted identically to sons. But why should disappear the autosomal signature of the other 22 chromosomes transmitted from father to children !?
 
Y chromosome is transmitted identically to sons. But why should disappear the autosomal signature of the other 22 chromosomes transmitted from father to children !?

It happens when, as an example, a woman or man of Amerindian or black ancestry mates with a person of European ancestry, and all of his or her descendants do as well.

With every generation each parent's contribution is cut in half, yes? So, Each person has 2 parents, 4 grandparents, 8 great grandparents, 16 great great grandparents, 32 great great great grandparents, and 64 great great great great grandparents. Someone could wind up with one Amerindian or SSA ancestor out of 64.

That would make that person l.6 percent of that ancestry. Yet, if the woman carried an Amerindian mtdna, or the man an Amerindian yDna, that could still show up.

In another two generations the autosomal signature for that ancestry could totally disappear in the descendants if they also marry people of European descent. That could happen in 200 years.

This is what happened to American politician Elizabeth Warren. Going by the fact that she's about 1-2% "Native American", she had one Native American ancestor about six to seven generations ago. That's one ancestor among 64 ancestors with that kind of ancestry. She used that tiny amount of ancestry to claim she was "Amerindian", because you get preferential treatment in getting jobs, getting into certain universities etc. She even put it on her Bar Association application (a lawyers' organization). It's totally dishonest imo, even if she carries an Amerindian mtDna.

You can see it clearly when it's admixture between different "continental" groups, but when it's all Europeans, it's much more difficult to disentangle.
 
It happens when, as an example, a woman or man of Amerindian or black ancestry mates with a person of European ancestry, and all of his or her descendants do as well.

With every generation each parent's contribution is cut in half, yes? So, Each person has 2 parents, 4 grandparents, 8 great grandparents, 16 great great grandparents, 32 great great great grandparents, and 64 great great great great grandparents. Someone could wind up with one Amerindian or SSA ancestor out of 64.

That would make that person l.6 percent of that ancestry. Yet, if the woman carried an Amerindian mtdna, or the man an Amerindian yDna, that could still show up.

In another two generations the autosomal signature for that ancestry could totally disappear in the descendants if they also marry people of European descent. That could happen in 200 years.

This is what happened to American politician Elizabeth Warren. Going by the fact that she's about 1-2% "Native American", she had one Native American ancestor about six to seven generations ago. That's one ancestor among 64 ancestors with that kind of ancestry. She used that tiny amount of ancestry to claim she was "Amerindian", because you get preferential treatment in getting jobs, getting into certain universities etc. She even put it on her Bar Association application (a lawyers' organization). It's totally dishonest imo, even if she carries an Amerindian mtDna.

You can see it clearly when it's admixture between different "continental" groups, but when it's all Europeans, it's much more difficult to disentangle.
Ok But... I understand that this subject is about admixture, and not about dilution of genetic contribution of one person or a very small group mixing with a larger group.
 
Ok But... I understand that this subject is about admixture, and not about dilution of genetic contribution of one person or a very small group mixing with a larger group.

It is "precisely" about the dilution of genetic contribution of one person or one small group "through" admixture.

That person's or small group's genes will disappear through admixture.
 
It is "precisely" about the dilution of genetic contribution of one person or one small group "through" admixture.

That person's or small group's genes will disappear through admixture.
I thought the admixture was just about large size groups. but I think that a very small group leaves traces in the big group, even if not in all the members of the resulting group.
 
Hello, I am somewhat new to this subject although I have been reading for hours about the matter for the last two days and I was aware of haplogroups for many months. When I "set out" to better understand the genetic make-up of Georgians I was very often encountering a lot of discussions about the proportions of admixtures in certain ethnicities, like 20% ANE and 80% EEF in Georgians or theories like Georgians being mostly Caucasian Hunter Gatherers and EEF and an unending list of theories basically. But this question is not about Georgians specifically but about the meaning of admixtures. What are admixtures precisely? When I first got into this subject I had the impression that Y-DNA and mtDNA are the sole or at least primary defining factors when it comes to national genetics and phenotype, but as I am aware now autosomal DNA is a more important factor and it combines the genetics of all our ancestors instead of father's father's father's genes and mother's mother's mother's genes.

So are admixtures a group of certain subclades? A certain proportion of Y-DNA haplogroups+mtDNA haplogroups? I have seen some admixture maps on this website and a lot of them added up to more than 100%, so are admixtures pretty much arbitrary? You take one ancient genetic sample and compare them to other groups? Like the West Asian or Caucasian admixture for example, it said that Georgians had around 90% of it, and every other region had less than that. Does that mean that other regions' populations are discussed from the point of view of being similar to Georgians or is the Caucasian admixture a very real and objective fact? Because on the eurogenes website I have read through so many discussions and opinions that I feel mind****ed at this stage and the topic of population DNA seems to be pointless speculation. Also the mtDNA in pretty much every ethnic group is so diverse that I struggle to understand how you could attribute a certain haplogroup to a certain ethnicity or a group of ethnicities. Do admixtures have a better correlation with phenotypes or with actual genetic make-up of a group of people?
https://www.well.ox.ac.uk/~gav/admixture/2014-science-final/resources/FAQ.pdf
 
I thought the admixture was just about large size groups. but I think that a very small group leaves traces in the big group, even if not in all the members of the resulting group.

It depends on the size of the minority admixing group and the amount of time which has passed. Our earliest R1b V-88, which is the yDna present in some herding groups in Chad and northern Cameroon, was first found in Europe among hunter-gatherers. I'm not aware of any study showing any of that autosomal ancestry, as represented by the autosomal chromosomes, in the African herders. Given they were only a group men, and the distance and time traversed, perhaps recombination cut the alleles into such small pieces that the ancestry of those European hunter-gatherers can no longer be found in those autosomes.

This is particularly a problem, I think, when you're dealing not with such easily distinguishable genomes as those of WHG, EEF, CHG/Iran Neo, but relatively modern European ethnicities.

When Ralph and Coop did an IBD analysis of Europeans, they couldn't go back more than 4300 years. That was a while ago, so they might be able to today, and this is IBD analysis not ADMIXTURE, but these groups are too similar, I think, for ease of attribution. Also, there was, for example, no "Polish" gene flow into Italy even in 2300 B.C., but there might have been admixture from steppe based groups.

k1AJlo9.png
 
It depends on the size of the minority admixing group and the amount of time which has passed. Our earliest R1b V-88, which is the yDna present in some herding groups in Chad and northern Cameroon, was first found in Europe among hunter-gatherers. I'm not aware of any study showing any of that autosomal ancestry, as represented by the autosomal chromosomes, in the African herders. Given they were only a group men, and the distance and time traversed, perhaps recombination cut the alleles into such small pieces that the ancestry of those European hunter-gatherers can no longer be found in those autosomes.
If R1b V-88 has a reproductive advantage over the original African haplogroup, then a very small percentage of European bearers would be enough to explain this dilemma.
 
Uniparental markers like yDna do not have "necessarily" anything to do with autosomal signatures, although often they do. You have to look at each case individually.

Take, for example, Cameroon herders of Africa, who carry R1b but are autosomally West Africans. When a group of predominantly male migrants enters a region and mates with local females, their autosomal signature can disappear very quickly in their descendants.

This is what I don't get. You say autosomally West African, so it is possible to trace someone's exact ethnic ancestry going back many millennia in the past? Are there some kind of small genetic mutations that are attributed to certain ethnicities? How exactly is ethnicity identifiable when you observe the genome.
 
Irakly,
However, we all are over 99% geneticaly identical.
But there are some mutations that are more numerous in one population than in another, or differs from one population to another. But I do not think there is a population that has all the genome mutation set, of its own or different of all others. So nobody belongs to a single pure "ethnicity" because it does not exist genetically. Everything is very relative. And we're joining a network.
The more we look back in time, over 1000 years, the more we know how mixed our DNA is.
 
I was expecting something more serious on this website than "we all come from Africa anyway dude" reasoning. We are also 96% genetically the same as Chimps, that doesn't make us 96% Chimps, does it? We aren't half bananas either. So every tenth of a percent is a very significant difference. Also my question was about how we differentiate between ethnicities on a genetic level. People can be mixed but they are always mixed only within a region. They are not usually 33% Negroid 33% Mongoloid and 33% Caucasoid. Different proportions of different-looking ancestors result in different phenotypes, and since there are different phenotypes, there are different genes, so I was asking how are the genes different between people with differing phenotypes. Do they have different alleles and such.
 
This is what I don't get. You say autosomally West African, so it is possible to trace someone's exact ethnic ancestry going back many millennia in the past? Are there some kind of small genetic mutations that are attributed to certain ethnicities? How exactly is ethnicity identifiable when you observe the genome.

People have different snps or alleles. Certain alleles are more prevalent in certain populations. They "cluster". Through ADMIXTURE programs, one can see what proportion of alleles from which ancestral group you carry.
 
As a scientific illiterate, and therefore not at all qualified to pronounce on these matters, I must say that I am most "skeptical" of autosomal findings. Y haplogroups and mtDNA are for me "hard" findings, whereas the autosomal readings vary from test to test to test. And so I must second Irakli's fundamental question, where the hell do the testing companies and on-line calculators get the autosomal numbers from? It all seems half-baked and artificial
 
As a scientific illiterate, and therefore not at all qualified to pronounce on these matters, I must say that I am most "skeptical" of autosomal findings. Y haplogroups and mtDNA are for me "hard" findings, whereas the autosomal readings vary from test to test to test. And so I must second Irakli's fundamental question, where the hell do the testing companies and on-line calculators get the autosomal numbers from? It all seems half-baked and artificial

The data is always the same. Only the interpretations differ. That may become better with time.

As for yDna and mtDna, again, the data is the data. The conclusions drawn from that data are interpretation and can vary, and what uniparental markers can predict about ancestry can be completely wrong. That's why you can wind up with E1a in a genetic and genealogical "native" Brit. That's also why a 100% Italian like me can have U2e as a mtDna. How is that helpful in telling me about the majority of my ancestors? The answer is that it's not helpful at all.

The people who complain about autosomal dna are usually the people who don't like what it has to say, predictably enough. :)
 
As a scientific illiterate, and therefore not at all qualified to pronounce on these matters, I must say that I am most "skeptical" of autosomal findings. Y haplogroups and mtDNA are for me "hard" findings, whereas the autosomal readings vary from test to test to test. And so I must second Irakli's fundamental question, where the hell do the testing companies and on-line calculators get the autosomal numbers from? It all seems half-baked and artificial

That's a matter of still developing technology in an extremely recente technology (precise, genome-wide and cheap autosomal DNA analysis is barely 10 years old), not a problem of the autosomal DNA data obtained per se. The Technologies (softwares and so on) devised to analyze and classify that huge amount of data may not be perfect yet, but they've improved tremendously in less than 5 years, and if not flawless they at least do give us a much clearer picture and at least approach the truth. Science is like that, it slowly builds up its knowledge to become closer and closer to the truth.

As for Y-DNA and Mt-DNA haplogroups, they're helpful, but also misleading. They tell you literally about 2 - the two "unbroken" ones - lineages out of hundreds of lineages that contributed to your ancestry. How is that necessarily representative of your overall ancestry? There are many populations out there that were formed after lots of gender-biased admixture events, so there is no necessary congruence between the Y-DNA and Mt-DNA distribution and their overall ancestry. There are I1 black men in Brazil and the USA, but that does not make them even necessarily "more European" than other black men carrying E1a or E1b1a who may actually have even more non-African ancestry than them.
 

This thread has been viewed 9733 times.

Back
Top