GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2004-05 > 1083466313


From: Bonnie Schrack <>
Subject: Re: [DNA] Questions about mtdna results requested
Date: Sat, 01 May 2004 22:51:53 -0400


Ann wrote in response to this mtDNA question:

>> Is there a table that describes the mutations and their probably causes with
>> the timeline in which it occurred?
>
> Yes, mutations are cumulative, so theoretically one could trace the
> sequence of events, if there were enough samples.

>Now look at the top, on the branch that leads to I1. Everyone has mutations
>at 203, 16172, and 16311. But only some have mutations at 16558, 322, and 295,
>so we conclude that those mutations came later. We can't tell the order,
>however. There is also a person with mutations at 16558 AND 10, so we conclude that
>10 occurred more recently. We can't tell the order of 203, 16172, and 16311,
>either, because we don't have any samples where one of those is missing.

Overall, Ann's explanation is excellent, but I'd just like to clarify a little bit more. The diagram she referred to was made by reading the complete mtDNA sequences of a sample of Finns. Of course the scientists already knew a good deal about the relationships between the haplogroups, but they used the specific data from these samples to create the specific branches of this diagram.

If they had used a different or larger sample set, it would have been possible to show many more branches. These branches, of course, do exist, but didn't happen to be present among those test subjects.

Since I study the I haplogroup, I can say with some confidence that 16311 probably occurred before 16172, because one can find many haplotypes with 16311 and without 16172, in our haplogroup. This can be seen in many data sets, such as in the "Founders" study, and among the members whose haplotypes are displayed on our web page: http://www.ancientrootsresearch.com/Hap-I/Members.html
Out of 20 haplotypes on that page, nine people have 16311; of those nine, six have 16172. So 16311 occurs three times without 16172, while 16172 doesn't occur without 16311 (almost never, in this haplogroup). This isn't an adequate sample by itself, but there's a lot more data out there that shows the same pattern.

So, likewise in any haplogroup. If you collect enough haplotypes, as John Walden has been doing, you can figure out to some extent (ideally, of course, one would benefit from formal genetic studies in order to do an optimum job of it) the order in which the mutations probably occurred. But putting a date on those mutations is extremely difficult with mtDNA. There is just not enough known yet to allow a clear and consistent rule to be applied the way it is with Y chromosome DNA. They are working on the problems involved -- it is slowly being hashed out. By reading scientific papers you can find various estimates.

There are many other diagrams out there, although Finnila's has been the best for a while. A new one has just been published, though, that I'd like to recommend ( I know I'm not the first one to mention it), especially since it includes haplogroups from all parts of the world, not only Europe. It's in the last paper listed on the FTDNA library page:
Lluis Quintana-Murci et al, "Where West Meets East: The Complex mtDNA Landscape of the Southwest and Central Asian Corridor."
http://www.ftdna.com/pdf/AJHG2004.pdf

The tree diagram is found on pg. 5. It's a very good and up-to-date phylogenetic tree for mtDNA. The tree is based on "the complete mtDNA RFLP haplotypes and HVS-I sequence data of 208 individuals from the Caucasus region, Anatolia, and southeast Pakistan." The whole paper is first-rate and well worth reading. New discoveries, new haplogroups, maps, etc., etc. You will find that the tree includes both Coding Region and HVS-I (HVR1) mutations, but not every mutation found in each haplogroup -- only the ones they found to be most essential in defining it.

The exciting thing it says in the paper is that their complete mtDNA data for all 910 samples is available in the online-only material! But so far, I don't know if that has been put online, or not. Does anyone have access to the Am. J. of Human Genetics without having to wait 6 months? I'm dying to see that data! If it doesn't show up, we'll have to write to Dr. Quintana-Murci, I suppose, or else inquire via FTDNA.

Bonnie Schrack




This thread: