MTDNA results mean???

[nathanm]

You nailed it perfectly! Would it make more sende to do the FMS on my kit or on my mothers kit?

Thanks

She literally co-wrote the book: Trace Your Roots with DNA, by Megan Smolenyak & Ann Turner.

Barring a mutation between your mother and you, the results should be identical. But in the highly unlikely scenario there was a mutation, it would be better to use your mother's kit.

[aeduna]

And I have exactly zero.

ditto, both for Y & mtDNA that I've done / arranged

[gtc]

ditto, both for Y & mtDNA that I've done / arranged

Well, for the time being anyway and with the current database, you can consider yourself unique.

[irshgypsy]

MtDNA

IMHO, the best way to go is the FF test and look for matches there to discover a common relative. It seems there needs to be a lot more research done before MtDNA is genealogically helpful. But remember--I'm an amateur!

MJ

[nathanm]

IMHO, the best way to go is the FF test and look for matches there to discover a common relative. It seems there needs to be a lot more research done before MtDNA is genealogically helpful. But remember--I'm an amateur!

MJ

The research on mtDNA is more advanced than autosomal DNA matching at present. The problem is mtDNA mutates slowly, so even an exact FGS match could mean a common ancestor hundreds of years back. According to the FAQ, it means a 50% chance of a common ancestor within 5 generations. Then it says 5 generations is about 125 years, which is on the low end of average generation length.

The "Understanding Your Results" PDF goes further, saying a 75% probability within 9 generations, and 90% within 16. If we use a more reasonable average generation length of 30 years, we're talking 480 years ago! Plus, there's still a 10% chance you don't have a common ancestor within 16 generations.

Some people get lucky, and a rare haplotype "confirms" a speculative relationship at HVR1 or HVR2. Others in more common haplogroups could have multiple matches at the FGS level, and not have any leads whatsoever to find a common ancestor.

[thetick]

Wow after seeing the percentages with an exact FGS match, I don't think I'll be buying the mtDNA FGS until it's around $100. I'm also factoring in the small number of FGS samples at FTDNA and Genbank for my known (23andme tested) subcalde , H5a1. $289 is far too much to spend for such a slim chance.

[marcusbgray]

And I have exactly zero.

Same here.

[JohnG]

I do not have any FGS matches listed, BUT if I look at the origins page, I find I have two exact matches that may have chosen not to be listed by email address. My oldest known maternal ancestor lived in Denmark. The matches were from the Netherlands and Germany. So I suspect travel along the coast of the North Sea, but don't know.

[gtc]

Wow after seeing the percentages with an exact FGS match, I don't think I'll be buying the mtDNA FGS until it's around $100.

I bought FGS when it was on special, and I really only did so to round out my DNA "profile" as tracing my maternal line is quite difficult due to female surnames being omitted from BD&M certificates.

As it turns out my mtDNA hg is quite rare, so it didn't really help ... other than to add a new subclade to the phylotree.

[RondaMiller]

I have over 8800 matches for the HVR1 none for HVR2 or the FGS lol....

[JohnLloydScharf]

Mutation Rates, Subclades, HVR1+HVR2, "Match"

You can establish a mutation with a few hundred thousand sperm. You cannot do that with ova. Even though sperm shed their mitochondria, I suspect that could be used to establish a mutation rate. I am sure the selection process affects the outcomes.

However, you cannot depend on who gets tested with a FGS. That is a non-random process with far too few samples. You certainly cannot have 100,000 babies to establish mustation within a haplogroup. If they have another way to establish mutation rates, I am happy to listen.

Regarding subclades of haplogroups, no distinction can be made between polymorphisms that define a subclade and those that do not.

The hypervariable region is called that for a reason. There is no need to keep it totally static to function. So, it is of little help in establishing a subclade.

The control region does not change fast because of repair mechanisms and mitochondria antibodies. You need those genes to function. Since the revised Cambridge Reference Sequence has rare polymorphisms, those are useless in discrimination of a subclade.

A subclade is established is if two or more Full Genome Sequences [FGS] have a polymorphism. That is a subjective rather than objective process. So, what is "private" or is used for a subclade is a moot point.

A mitochondria from a cell [like the liver] can number over a 1000. There are always mutations within a cell. Some are repaired. Some are killed by antibodies. As you get older, the repair/kill function becomes less effective. So, "heteroplasmy" is always exists. If it reaches 30%, it is reported.

Without a FGS, you are only guessing. I have two FGS matches and 1300 who have matched me with HVR1+HVR2. Of those 1300 matches, over 200 have had a FGS. That means I have a one in 100 chance of matching plus or minus 1.

I suspect out of that 1100 who have NOT tested their FGS, that I may have another ten FGS matches, using that swag factor [Scientific Wild * Guess]. You may have a 50% chance of a match being from an ancestor with a FGS, but you can exclude with certainty anyone who does not have a high resolution match.

FTDNA's FAQ gives a 50% probability for a common ancestor within 5 generations for a perfect FMS match.

http://www.familytreedna.com/faq/answers/default.aspx?faqid=10#486

However, they haven't explained the methodology they used to arrive at that figure, and I believe it is over-optimistic. Not that I have an alternative to offer -- I don't think we understand enough about mtDNA mutation rates. What I generally tell people is sort of a truism: sharing a number of "private" mutations (ones not used to define haplogroups) pulls the MRCA toward the present.