GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2011-03 > 1299871325


From: SVass <>
Subject: Re: [DNA] Interesting mtDNA example
Date: Fri, 11 Mar 2011 11:22:05 -0800
References: <mailman.345.1299830449.7803.genealogy-dna@rootsweb.com>
In-Reply-To: <mailman.345.1299830449.7803.genealogy-dna@rootsweb.com>


I have a similar example within the H1e haplogroup. There is an Ashkenazi cluster with approximately 24 HVR1/2 matches that has six exact FGS matches and three more with one extra coding region mutation each. There is a Ashkenazi second cluster of three FGS matches that have the same extra mutation in HVR1 (with sixteen matches in HVR1/2). In other words, FTDNA software did not intelligently search for possible clusters as a human could do.

This is not a simple problem for any software designer to solve. If they provide an infinite number of possibilities, they will overwhelm the system and the average user.

For those who care about my version, the first H1e group has for HVR1, 16114 and 16519. The second has 16114, 16126, and 16519.
sam vass
R1b-V88+
H1e

On Mar 11, 2011, Aaron Salles Torres wrote:
> ...I manage the Fernandes DNA Surname Project, where I also allow the display of mtDNA results - as this is an Iberian surname, and in Iberia last names are often inherited matrilineally.
>
> If you take a look at our results section ( http://www.familytreedna.com/public/Fernandes/default.aspx?section=mtresults ), you will notice two individuals in the C1 haplogroup, 186136 and 190662. 186136 is 190662's son and they both have ordered the FGS test (not displayed due to privacy reasons). In any case, you will notice that the son has a mutation in the HVR2 panel (204Y) not presented by the mother. This is interesting in itself, because if we did not know they were related, we would assume there was just "50% chance the MRCA lived 1300 years ago."
>
> This is because - and here is what is more interesting to me - the mother does not see the son as a match on the HVR2 panel, nor under the FGS results. The son, however, does see the mother as a match under both HVR2 and FGS. The implication is that if your mtDNA line has remained unchanged while a close relative has had a mutation, you will not be aware that the individual is so close to you - you will not see them as a match under HVR2 or beyond.
>
> How often does this happen (a close relative having an "unexpected" mtDNA mutation like this) and how many matches are we missing in our genealogical explorations due to the way FT DNA displays these results?
>
> I already have had similar problems with Y DNA results, in which two close individuals have one mismatch in the 12 marker panel. As there are not many individuals outside of the US and Britain who have been tested by FT DNA, many international clients happen to have no matches under the 12 marker panel. That is discouraging and, for that reason, they do not order further tests. When I have caught such cases on ysearch and convinced such individuals to test more markers, they were a 65/67 or 66/67 match with someone whose profile I managed. This is what I call a false negative. So why not show ALL 11 marker matches, whether two individuals are part of the same project or not? Same thing seems to apply to mtDNA results: are the match displaying standards too strict? Shouldn't it be up to the clients to choose which matches they should pursue or not?
....

This thread: