Explain my mitochondrial DNA

July 14, 2005 12:47 PM

I just received my results from the Genographic Project and they are confusing. Can someone explain my mitochondrial DNA to me?

1. They said that I am part of group L1.
2. L1 is not found outside of Africa.
3. All existing groups descend from L1.

First off, I am not African at all. My first thought was that they came to that conclusion because of the Filippino and Cuban parts of my ethnicity which can have traces of L1, but that only goes back 3 generations before there's a roadblock. My great grandmother was French, so, L1 shouldn't be in my mtDNA since it came from the male members of my geneology, right?

Then I thought that I was determined to be L1 because maybe I have so many different branches, that the only thing that is certain is that they have a common ancestor in L1.



Looking at my mitochondrial diagram, it has a bunch of different letters that might correspond to the different threads (ND, CO, HVR?) that split off from L1 when it started migrating. But wouldn't that be recorded in my mtDNA's journey? Isn't that the point of this project?

Geneticists, help! I've found some blogs who have posted their results, which all seem markedly different than mine, at least travelling out of Africa.

posted by scazza to Science & Nature (14 answers total)

Remember, mtDNA is only inherited in the egg, from females. So your mom, your mom's mom, and so on, in a single long chain. Not a branching family tree.

Remember how it works: when cells divide mitochondria (which have their own DNA) get split between the two cells, just like the nuclear DNA. But in Meiosis, extra DNA comes from the sperm, but all the cellular internals, including the mitochondria come from the egg cell when it starts dividing.

<speculation>
Although, hmm, I suppose some of the mitochondria from the sperm could get into the egg, and be carried along with the rest of the cells? But that would mean you'd have several strains of mitochondria, rather then having mixed mtDNA like you would with your regular DNA. Hopefully someone who knows more about the process can chime in
</speculation>

Until then, try to think back on your mom's side. If a black woman had children with a white man, and her daughter had children with a white man, and so on for a couple generations, you'd look white but have African mitochondria, as far as I know
posted by delmoi at 1:23 PM on July 14, 2005

My great grandmother was French, so, L1 shouldn't be in my mtDNA since it came from the male members of my geneology, right?

I don't know how to read the data you have, but all of your mitochondrial genome comes from your mom. No mitochondrial genetic information comes from your father. This is true for everyone. This is why mitochondrial diseases are passed down from female carriers but not male carriers.
posted by Rothko at 1:27 PM on July 14, 2005

The cool thing about mt DNA is that, unlike human DNA which gets mixed and matched every generation, is that mtDNA stays constant, and only changes very slowly, and at a very specific rate. You can literally take two people and tell (almost) exactly how many years ago they split off from a single matrilineal ancestor.

As far as your total genetic history, well, at each generation backwards you have more and more ancestors. Eventually the size of a single generation of ancestors is as large as the entire human race at the time. Pretty much everyone is only a few generations removed from everyone on earth, I think.
posted by delmoi at 1:33 PM on July 14, 2005

Response by poster: Right, yes mtDNA comes from my mother's line and what I'm saying is that it is European. So where did they get L1 from? When you look at the other bloggers results, they at least get to Europe. The Indian guy got to India. I'd think that at least there would have been some migration in my mtDNA to Europe.

delmoi, it seems like the project is tracing migration through DNA, not just the origin of ancestors. Otherwise everyone would be L1. That is why this is confusing, why am I just L1?
posted by scazza at 1:46 PM on July 14, 2005

Why couldn't one of your french ancestors have been from Africa, and re-introduced the L1 mtDNA. Wasn't there a bit of African influence in France in the 1500s and 1600s? Or was that mostly Spain...
posted by dness2 at 2:11 PM on July 14, 2005

scazza, what delmoi is saying is that it might be possible that you've got a black ancestor whose daughter or granddaughter passed for white - I'm not sure how common slaves were in France proper, but I know that their Carribean colonies way back in the 17th and 18th centuries had a lot of slaves. Unfortunately, as delmoi says, when you trace mitochondrial DNA this way, you're pretty much following a single line through your family tree. This means that even if you had many many white French ancestors, a single black maternal ancestor in the right place on your family tree throws everything off, and you get told you're part of the L1 group. That's one of the reasons this technique has only limited value, since few people have homogenous family backgrounds, and this technique can only tell you one of the many places your ancestors came from [and the answer they give you isn't even necessarily the place the majority of your ancestors were from.]
posted by ubersturm at 2:17 PM on July 14, 2005

Right: you might be 1/256th black, and your mtDNA will still be african. Your great grandmother might have been born in the 1850s or something, and there were a few africans living in france at the time, and before then.
posted by delmoi at 2:30 PM on July 14, 2005

So basically, it's partly just luck that the bloggers with European and Indian ancestry happened to get back to the places where a majority of their ancestors were from. Either their families happen to have been more stationary and homogenous [certainly possible, particularly if their families immigrated more recently], or ancestors not from that background happened not to be in the line of direct maternal descent, and therefore didn't pass on their mitochondrial genes.
posted by ubersturm at 2:39 PM on July 14, 2005

Best answer: scazza, my husband I and I have both tested with FamilyTreeDNA, the company that is also handling the backend processing for the Genographic Project. His mtDNA is T1 and mine is H2*. I hope our stories help you understand why your results are, in fact, likely to be correct:

My husband's mother's family are Sephardic/Mizrahi/Romaniote Jews from the former Ottoman Empire--only we're not really sure what the exact admixture is, since the immigrant Sephardic (Spanish-Jewish) Jews intermarried with the existing-since-ancient-times Romaniote (Greek-Jewish) community, pretty much swamping them and dominating them with their population size, rituals, and language(s). And then they intermarried with the Mizrahi (Turkish-Jewish and Middle-East-Jewish) community. But it all gets referred to as "Sephardic", which isn't really quite accurate, though numerically and socially they dominated the region.

Enter genetic genealogy. Doing a mtDNA test showed that my husband is mtDNA haplogroup T1, which has origins in southern and Mediterranean Europe. Within FTDNA's proprietary database, he also has HVR1 matches (no HVR2 matches, alas) to two other people, both of whom report Syrian-Jewish or Greek-Jewish maternal ancestry. In this case, his mtDNA helped confirm that his mother's mother's...mother (direct female-only line) was a Romaniote or Mizrahi Jew, native to the area, not one of the Sephardic Jews who fled from Spain and Portugal to Rhodes between 1492-1522.

So, in his case, his mtDNA helped establish something we already suspected, since there was pattern of Sephardic immigrant men coming in to the area and intermarrying with the native Mizrahi and Romaniote women. (Incidentally, that social pattern persisted many centuries, where local men in Rhodes, usually with Sephardic surnames, would "send away" for brides to be brought in from the Turkish mainland. My husband's Turkey-born but Rhodes-married great-grandmother was one of them; her family surname originally came from Sfad, Palestine, but her husband's surname was Sephardic.)

In my case, my mtDNA haplogroup is H2*, H with all its subgroups being the most common maternal haplogroup in Europe and very very common in the UK. Furthermore, my HVR2 (high-resolution) matches were mainly from people with Scottish or Scots-Irish maternal ancestry.

This seriously confused me when I first got my results back, because my mother's mother's...mother's line can be traced back (with a paper trail, and everything) to an Ashkenazic Jewish woman born in Warsaw, Poland in 1800. And my entire family on all sides, going back many generations, is plain-vanilla Ashkenazic Jewish. So how and why would a Scotswoman marry into a Jewish family? Weren't conversions disallowed by her community? (And, um, weren't most Europeans burning us back then?) There weren't even any Ashkenazic Jews in Scotland in the middle ages, were there? (Sephardics, yes, but just a tiny number.)

Well, after a bit of historical digging, it turns out that there were tens of thousands of Scots in Poland, especially Warsaw, between the 16th-18th centuries, trading their wool. And there were thousands more Scots in the Baltics. That's because the Poles, at the time, were not going to persecute the Scots for their (non-Catholic) religious beliefs. So Scots traders meet up with Jewish traders in Warsaw, do a little business, and apparently love bloomed between at least one couple, with the woman converting to Judiasm, marrying a Jewish husband, and staying behind in Warsaw.

Or, at least, that's the current theory. So in this case, unlike my husband's, my mtDNA was quite a bit of a surprise.

Now, your L1 results, as others here have pointed out, are relevant only to your mother's mother's mother's...mother's line, your direct female line ONLY. They refer to only one person out of your entire ancestry pool. But L1 is an unmistakenly African haplogroup.

If you want to, you can test with a company called AfricanAncestry.com that specializes in telling you exactly which region (sometimes even exactly which tribe!) your ancestry is from in Africa.

By the way, please add your results to the FamilyTreeDNA pool--there should be a link somewhere at the bottom of one of the Genographic Project pages to upload your results for free to the FTDNA database. It befeits everyone to have a larger pool of results available for comparison.

And you should totally bug your father or your brother, if you have one, to get their Y-DNA tested. It gives much more recent results than your mtDNA, because it mutates much much faster. Meaning, if his Y-DNA haplogroup is, say, E3b1a, you'll have a good localized region to look for ancestors. (Albania and Macedonia in that case, I think.) And if you find a close match with another person, like 35 out of 37 markers, you'll probably have a common ancestor between you within the past 400 years. Whereas people who have HVR1 and HVR2 matches on mtDNA only have ancestors in common sometime within the past 1000 years or maybe more.
posted by Asparagirl at 4:32 PM on July 14, 2005 [2 favorites]

Just a note about the Y-DNA--it only tests your father's father's father's...father's line, your direct paternal line. So, once again, you're only going to get information from it for only one branch of your family tree. But since that branch should also carry your surname, since surnames are usually passed from father-to-child, it makes doing paper-trail research quite a bit easier.

And anyone else who tested with the Genographic Project--not only can you upload your results (for free) to FamilyTreeDNA, but you can have them run discounted upgrade tests on your DNA sample, if you want higher-resolution results. For guys, this means you can take your 12 markers and upgrade them to 37.

Also, check out mitosearch.org and ysearch.org if you want to type in your Genographic results and search their database (for free) to see who you match so far, and where they came from.
posted by Asparagirl at 4:42 PM on July 14, 2005

Those of you that have done this -- is it worth the money?
posted by k8t at 7:48 PM on July 14, 2005

I don't know how to read the data you have, but all of your mitochondrial genome comes from your mom. No mitochondrial genetic information comes from your father. This is true for everyone.

It's only true to a first approximation. Cases exist where mtDNA from the father was found in the children, in varying percentages.

It's been hypothesized that if the mother had a mitochondrial disorder, and a few mitochondria entered with the sperm, that the paternal mitochondria, being more 'fit', could out-compete the maternal mitochondria enough to make up a significant fraction of the cell's mitochondria by the time of first mitosis.

But that's all handwaving. What you should know is that, although the above posters' dictum is almost always true, it is not inviolable dogma and exceptions have been recorded.
posted by ikkyu2 at 9:55 PM on July 14, 2005

Those of you that have done this -- is it worth the money?

If you're curious about knowing a bare-bones "where did my very distant ancestors live" kind of thing, I would say go for the Genographic Project, because it's cheap and kinda neat and has pretty graphics on their webpage.

If you have done a decent amount of regular genealogical work already, but are stuck at a real brick wall, or are running out of a paper trail because records were either destroyed or just don't exist beyond a certain point--that one was my motivation, since Ukrainian records are spotty--then I say go for FamilyTreeDNA.com. But it's not terribly cheap; a 25-marker Y-chromosome DNA test (which only guys can take) is a little over $210, I think. They can take the 12-marker test instead, which is less expensive, but it's too few markers to really be useful, IMHO. And both men and women can take the mtDNA test. And there are neat surname studies going on to reconnect people with common names, like Rose, Smith, Donald, whatever, into a recognizable family tree. I started up one for my husband's surname, but alas, he's the only one in it so far.

AfricanAncestry.com is good if you already know you have African roots and want to mine their specialized database of African results. And I've also heard good things about Relative Genetics and Oxford Ancestors, and a few other firms. The prices are all pretty similar.

But remember, genetic genealogy is meant to be an adjunct to regular genealogy, not to totally supplant it. You can research census records and land records and even some birth and death records and such at any LDS Family History Center or local library for free, and that should be your first stop before deciding whether to take the next step and send some cheek cells and a check off to a lab at the University of Arizona.
posted by Asparagirl at 11:34 PM on July 14, 2005

Response by poster: k8t, I did this in particular because I wanted to be part of it as a project, not necessarilly for the personal gain. I was interested if my ancestry was by route of Asia or something crazy like that. And it is something crazy like that. There are more expensive test services I think, that will give you more elaborate results, so I'd say that worth depends on what you are looking for. I'm satisfied, if confused, as it has given me a very interesting mystery to solve.

Thank you everyone, this has been interesting. Especially since, "[Y-DNA] gives much more recent results than your mtDNA, because it mutates much much faster." Most of those bloggers I cited are men, so of course they'd have more thorough results. North Africans are common in France, but I will have to research West and Central Sub Saharan African diaspora.
posted by scazza at 10:24 AM on July 15, 2005