Conflict between Autosomal and M-DNA.

[JFWR]

A curious turn of events has placed my M-DNA and maternal autosomal test results entirely in conflict with one another.

My M-DNA haplogroup is h-64, and every hit or near hit is concentrated almost exclusively in the British Isles (especially Ireland) or Americans with Irish ancestors. Given that my lineage on my maternal line is historically Irish, this was not at all surprising whatsoever. Indeed, I have pretty good geneaological evidence into the early 19th century that my maternal line is Irish.

Then I went ahead and did an autosomal DNA test on my mother owing to the weird results I was getting on my own autosomal test. As it turns out, my maternal autosomal test has come back with...

Nothing Irish at all, nor German (which is my grandmother's father's side).

How can this be? On the mitochondrial side, I am seeing a ton of Irish, including one person who is almost certainly a distant cousin owing to her genetic difference of 0 at full M-DNA sequence. But the autosomal tests give -no credit- to this.

I've compared my kits to my mother's and father's autosomal kits and have come up as their son on both accounts. But then my breakdown on Family Tree doesn't even match the breakdown the genetics would imply:

Father: 73 British, 25% Westeran/Central Europe, 3% Scandinavia.


Mother: 51% Scandinavia, 21% Southern Europe, 10% Finland, 17% Jewish diaspora


Myself: 36.5 British, 12.5% CEntral Europe, 26.5% Scandinavia, 11.5% Southern Europe, 5% Finland, 9% Jewish dispora by my approximate calculations.

Actual results: According to FT: Scandinavian 32%, Western and Central Europe 25%, British isles 19%, Finland and Northern Siberia 5%, Southern Europe 4%, Jewish diaspora 14%.

I don't see how this makes sense at all.

So I am left very, very much confused. Is this reflective of the poor accuracy of autosomal tests? Or what could be going on here?

I am most concerned about the obvious contradiction between a maternal haplogroup that speaks to definitive Irish origins, but autosomal data that suggests no such links, and indeed, links to far flung places for which I have no geneaological connection that I know of.

 

[Angela]

A curious turn of events has placed my M-DNA and maternal autosomal test results entirely in conflict with one another.

My M-DNA haplogroup is h-64, and every hit or near hit is concentrated almost exclusively in the British Isles (especially Ireland) or Americans with Irish ancestors. Given that my lineage on my maternal line is historically Irish, this was not at all surprising whatsoever. Indeed, I have pretty good geneaological evidence into the early 19th century that my maternal line is Irish.

Then I went ahead and did an autosomal DNA test on my mother owing to the weird results I was getting on my own autosomal test. As it turns out, my maternal autosomal test has come back with...

Nothing Irish at all, nor German (which is my grandmother's father's side).

How can this be? On the mitochondrial side, I am seeing a ton of Irish, including one person who is almost certainly a distant cousin owing to her genetic difference of 0 at full M-DNA sequence. But the autosomal tests give -no credit- to this.

I've compared my kits to my mother's and father's autosomal kits and have come up as their son on both accounts. But then my breakdown on Family Tree doesn't even match the breakdown the genetics would imply:

Father: 73 British, 25% Westeran/Central Europe, 3% Scandinavia.


Mother: 51% Scandinavia, 21% Southern Europe, 10% Finland, 17% Jewish diaspora


Myself: 36.5 British, 12.5% CEntral Europe, 26.5% Scandinavia, 11.5% Southern Europe, 5% Finland, 9% Jewish dispora by my approximate calculations.

Actual results: According to FT: Scandinavian 32%, Western and Central Europe 25%, British isles 19%, Finland and Northern Siberia 5%, Southern Europe 4%, Jewish diaspora 14%.

I don't see how this makes sense at all.

So I am left very, very much confused. Is this reflective of the poor accuracy of autosomal tests? Or what could be going on here?

I am most concerned about the obvious contradiction between a maternal haplogroup that speaks to definitive Irish origins, but autosomal data that suggests no such links, and indeed, links to far flung places for which I have no geneaological connection that I know of.

There is no conflict.

To put it simply, your mtDna comes from one ancestor among thousands. Your autosomal dna includes the dna of all of your ancestors.

On the autosomal side, Scandinavians could be categorized as a "Germanic" people.

I'm not particularly fond of the FTDNA autosomal analysis, but none of the commercial analyses are going to be able to give your the specificity you're looking for.

As to your mtDna, there may be a lot of people you're matching on it at a certain level of analysis who happen to come from the British Isles. Have you had a full mitogenome analysis? If you haven't, you might find that your matches would dwindle in number. Anyway, the British Isles doesn't "own" this particular mtDna. You have to remember that most of the people who test with these companies are from the British Isles or former British colonies. Academic studies which assemble samples from other places and in a randomized way present a different picture.

For H24, this is what is listed in our Genetics analysis on this site.

"H24: found in Britain, Ireland, Spain, Germany, the Czech Republic, Poland, Denmark and Finland / found in the Funnelbeaker culutre in Sweden."

You might find the whole article, and indeed the whole Genetics section of this site interesting reading:
http://www.eupedia.com/europe/Haplogroup_H_mtDNA.shtml

If you want to know where your mtDna line lived before coming to the U.S. you'd really have to do a genealogical search through records.

 

[Fire Haired14]

My Dad has H64, his H64 is from Norway. Where is your maternal line/H64 from?

 

[Angela]

There is no conflict.

To put it simply, your mtDna comes from one ancestor among thousands. Your autosomal dna includes the dna of all of your ancestors.

On the autosomal side, Scandinavians could be categorized as a "Germanic" people.

I'm not particularly fond of the FTDNA autosomal analysis, but none of the commercial analyses are going to be able to give your the specificity you're looking for.

As to your mtDna, there may be a lot of people you're matching on it at a certain level of analysis who happen to come from the British Isles. Have you had a full mitogenome analysis? If you haven't, you might find that your matches would dwindle in number. Anyway, the British Isles doesn't "own" this particular mtDna. You have to remember that most of the people who test with these companies are from the British Isles or former British colonies. Academic studies which assemble samples from other places and in a randomized way present a different picture.

For H24, this is what is listed in our Genetics analysis on this site.

"H24: found in Britain, Ireland, Spain, Germany, the Czech Republic, Poland, Denmark and Finland / found in the Funnelbeaker culutre in Sweden."

You might find the whole article, and indeed the whole Genetics section of this site interesting reading:
http://www.eupedia.com/europe/Haplogroup_H_mtDNA.shtml

If you want to know where your mtDna line lived before coming to the U.S. you'd really have to do a genealogical search through records.

Sorry, I thought you said H24. The data says H64 is found in Norway. There may be another cluster in the British Isles which hasn't been picked up by the academic studies.

 

[Maciamo]

As Angela said, it is perfectly normal to have different origins for the Y-DNA, mtDNA and autosomal results. And it is true that FTDNA My Origins is far from accurate in determining a person's autosomal composition.

If you use autosomal calculators like those found on GEDMatch (upload your genome first), you will be able to compare your genetic make-up with that of other individuals in all European countries. I made autosomal maps based on such calculators (mostly Dodecad K12 and K12b). This will not tell you where your recent ancestors come from, but rather what your overall ancestral admixture looks like, and in which country people carry a similar (ancient) ancestral admixture.

I would not advise you to order the mtDNA full sequence. This is generally useless and overpriced, because anyway mtDNA cannot tell you about recent or even not so recent ancestors. Most mtDNA deep clades already existed during the Neolithic or Bronze Age. So at best you'll know where you ancestors came from 5,000 to 10,000 years ago. If you have money to spend, I'd rather recommend a full Y-DNA sequence (Big Y if you are already with FTDNA or otherwise FGC's Y Elite 2.1), as this can give you an accuracy of just a few generations ago, and you will know the whole history of your patrilineal ancestors.

In any case, I strongly encourage you to read the article Which DNA test should I choose?

 

[JFWR]

Sorry, I thought you said H24. The data says H64 is found in Norway. There may be another cluster in the British Isles which hasn't been picked up by the academic studies.

The problem may also have to do with the fact thta H64 appears to be quite rare. It is barely mentioned in any references to Haplogroup H that come about. Through the "family finder" thing at family tree, I get only two people in Norway, but a bunch of people in the US and the British Isles.

Given that my ancestry by geneaology suggests a Dublin extraction, could it be the Norwegian lineage might express the Viking settlement of Dublin?

Thank you very much for you (and everyone else's) responses, by the way. I am highly appreciative of your assistance with this.

 

[JFWR]

My Dad has H64, his H64 is from Norway. Where is your maternal line/H64 from?

My genealogical ancestry suggests that my maternal DNA is Irish for at least the turn of the 19th century, if no deeper. So I am thinking perhaps the Norwegian ancestry comes by way of the Viking settlement of Dublin, as I hypothesized above to Angela.

 

[JFWR]

As Angela said, it is perfectly normal to have different origins for the Y-DNA, mtDNA and autosomal results. And it is true that FTDNA My Origins is far from accurate in determining a person's autosomal composition.

If you use autosomal calculators like those found on GEDMatch (upload your genome first), you will be able to compare your genetic make-up with that of other individuals in all European countries. I made autosomal maps based on such calculators (mostly Dodecad K12 and K12b). This will not tell you where your recent ancestors come from, but rather what your overall ancestral admixture looks like, and in which country people carry a similar (ancient) ancestral admixture.

I would not advise you to order the mtDNA full sequence. This is generally useless and overpriced, because anyway mtDNA cannot tell you about recent or even not so recent ancestors. Most mtDNA deep clades already existed during the Neolithic or Bronze Age. So at best you'll know where you ancestors came from 5,000 to 10,000 years ago. If you have money to spend, I'd rather recommend a full Y-DNA sequence (Big Y if you are already with FTDNA or otherwise FGC's Y Elite 2.1), as this can give you an accuracy of just a few generations ago, and you will know the whole history of your patrilineal ancestors.

In any case, I strongly encourage you to read the article Which DNA test should I choose?

Maciamo, as per usual you are such a helpful fellow. Sincerely, thanks for your help.

So, let me get this straight: Autosomal DNA tests don't test for relatively recent DNA, but rather which countries match something close to your mix of ancestral components? Given that they give modern countries in most of these analyses, I had assumed that this was closer to modern populations and reflective of genetic markers associated with those.

I took from your maps that effectively autosomal DNA tests reflect fairly ancient DNA comparisons, such that the modern populations are those which match nearest to the percentage of ancestry you have to these ancient populations. Is that how I should be reading them?

You see, what is curious is that two elements of my ancestry don't appear to be showing up at all: Evidence of Irish or German extraction, which is efffectively what my grandmother was (with the possibility of English as well). That is why I have been quite confused. What might that entail? That my ancestors from these locations were in other parts of the world over 500 some years ago?

Or basically: What exactly does it mean when they say "you're _____" on an autosomal test? Does it mean that the populations match some ancestral mix, or that you're testing compared to discovered genotypes of old bodies, or what precisely?

 

[New Englander]

You ever heard of Vikings? The moved from Scandinavian nations to England and Ireland a lot....Your first ancestor just happened to be a native of Ireland, and the rest was diluted.

 

[MOESAN]

senseful answer Angela - (private states about haplo's are very confusing as they are so biased) -
the differences between real autosomes test and waited autosomes %s from parents are not neglictible here, but not impossible -
we are not the "average" of our parental autosomoms but a cocktail of genes choosed among the half of them (I do simple here) 1/2 from father line, 1/2 from mother line - but every parent give us 1 chromosom out of 2 , not a gene out of 2: it makes some difference: spite some crossing-overs between homologous chromosoms during gametes maturation, the choice of a chromosom against another is not the same as a 1/2 choice between sand grains submitted to pure hazard; it results sometimes some visible differences between the verified auDNA of children compared to parents auDNA mean even if it never reaches great differences (because there are 22 pairs of chromosoms and not a lone one) but the most often the differences are very tiny -

 

[Maciamo]

Maciamo, as per usual you are such a helpful fellow. Sincerely, thanks for your help.

So, let me get this straight: Autosomal DNA tests don't test for relatively recent DNA, but rather which countries match something close to your mix of ancestral components? Given that they give modern countries in most of these analyses, I had assumed that this was closer to modern populations and reflective of genetic markers associated with those.

I took from your maps that effectively autosomal DNA tests reflect fairly ancient DNA comparisons, such that the modern populations are those which match nearest to the percentage of ancestry you have to these ancient populations. Is that how I should be reading them?

You see, what is curious is that two elements of my ancestry don't appear to be showing up at all: Evidence of Irish or German extraction, which is efffectively what my grandmother was (with the possibility of English as well). That is why I have been quite confused. What might that entail? That my ancestors from these locations were in other parts of the world over 500 some years ago?

Or basically: What exactly does it mean when they say "you're _____" on an autosomal test? Does it mean that the populations match some ancestral mix, or that you're testing compared to discovered genotypes of old bodies, or what precisely?

Actually it is possible to test recent ancestry with autosomal DNA. That's even the best way to do it. But that is what 23andMe calls DNA Relatives, while FTDNA has Family Finder. These software will analyse segments of chromosomes that you share with other individuals. Based on the number and length of these segments, they estimate how closely related you are to another person (parent-child, sibling, 1st cousin, 2nd cousin, etc.). However I wouldn't trust predictions beyond 2nd cousin. I know all my 1st, 2nd cousins, and know the name of at least my 3rd and most of my 4th cousins. Yet 23andMe returns matches of potential 3rd and 4th cousins in countries where it is impossible for me to have such close relatives, and whose surnames do not match at all those of my family. I contacted several of them to be sure, and there is no possible connection. So be careful. Anyway I don't think it is useful to think of 3rd or more distant cousins as relatives as they don't share more DNA with you than any stranger from the same ethnic background.

The autosomal reports you mentioned above, with the percentages by country, only look at more distant ancestry (say 1,000 to 2,000 years ago). For example I noticed that French-speaking Belgians generally scored 0 to 4% of Italian at 23andMe, while Dutch-speaking Belgians had between 0 and 1.5%. I think this reflects ancient Roman ancestry (at least part of it as it surely doesn't match exactly and even modern Italians sometimes score only 30% Italian).

 

[JFWR]

Actually it is possible to test recent ancestry with autosomal DNA. That's even the best way to do it. But that is what 23andMe calls DNA Relatives, while FTDNA has Family Finder. These software will analyse segments of chromosomes that you share with other individuals. Based on the number and length of these segments, they estimate how closely related you are to another person (parent-child, sibling, 1st cousin, 2nd cousin, etc.). However I wouldn't trust predictions beyond 2nd cousin. I know all my 1st, 2nd cousins, and know the name of at least my 3rd and most of my 4th cousins. Yet 23andMe returns matches of potential 3rd and 4th cousins in countries where it is impossible for me to have such close relatives, and whose surnames do not match at all those of my family. I contacted several of them to be sure, and there is no possible connection. So be careful. Anyway I don't think it is useful to think of 3rd or more distant cousins as relatives as they don't share more DNA with you than any stranger from the same ethnic background.

The autosomal reports you mentioned above, with the percentages by country, only look at more distant ancestry (say 1,000 to 2,000 years ago). For example I noticed that French-speaking Belgians generally scored 0 to 4% of Italian at 23andMe, while Dutch-speaking Belgians had between 0 and 1.5%. I think this reflects ancient Roman ancestry (at least part of it as it surely doesn't match exactly and even modern Italians sometimes score only 30% Italian).

Okay.

So basically, the family finder reflects actual, living relatives, whereas the ethnic percentages reflect a 1000-2000 years previous?

So if it says you're "Italian", it means your ancestors may or may not have been Italian 1000 years ago? What do they base this off of? DNA analysis off corpses dated by archaeological means?