GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2003-11 > 1067721248


From: David Faux <>
Subject: Re: [DNA] Marker Mutation Rates
Date: Sat, 1 Nov 2003 13:14:10 -0800 (PST)
In-Reply-To: <5.0.2.1.2.20031101090758.02583ba0@mindspring.com>


Hello Ron and all:

I fear that the "MRCA" (0 to 800 years or some such figure to the most recent common ancestor) concept and it's underpinning via "accepted" mutation rates (e.g., .002 per marker per transmission or whatever) is deeply woven into the thinking of most of us on the List. Actually I am of the opinion that the whole conceptualization is flawed - and totally agree with you. As a matter of fact I don't even give it a passing thought in my surname and geographical surname studies.

It is all guesswork, which can result in many false positives (assumption that people share a common ancestor in a genealogically meaningful time frame) and false negatives (assuming that two people are unrelated since there is one too many mutation such as only a 22 of 25 marker match). Studies of father to son transmissions indicate that mutations can occur at any time, even two step mutations. This one observation creates chaos in the concept of MRCA - which is based on probabilities and mathmatical models. Models schmodels. Mutations can occur two at a time (how can you known if it was 10 years ago or 1000 years ago - except via extensive collection of data from as many people with the same surname as is possible?). No one knows how many mutations are "back mutations". No one can say with authority that just because two people match at 24/25 that it is not merely coincidence and/or an example of convergent mutation. Cousins sometimes "mismatch" by 3 mutations - wh!
at
happens to their lines 400 years from now. Probably more mutations and the conclusion that there is no relationship. The MRCA world is one of extreme chaos - at least that is my conclusion. Even in the best case scenario of decision making re whether two people are related, and when, the plus or minus x number of years factor is so huge that at this point I have dispensed with the matter entirely.

In my case I have a distant "cousin" who is a one mismatch on a "slow moving" marker, and a two step mismatch on a "fast moving" marker. My only interest in the matter is to locate more family members to see if his "signature" is closer to our common ancestor of 400 years ago, or whether my line has better preserved the "original" haplotype. None of the calculators or estimates is of any help whatsoever. A strong stance. Perhaps, but the statistical MCRA estimates available are no better than guesswork, but have the aura of "science" surrounding them.

I hope that someone will step forward and offer a rebuttal to our take on this situation.

David F.

Ron Lindsay <> wrote:
Greetings Genealogy-DNA List,

Acknowledging the use of a small sample size (statistically a sin) to fan
the flames coupled with what appears to be "factual" genealogical detail, I
continue to look for legitimate reasons to challenge the current thinking
regarding the assumed Y-chromosome marker mutation rates.

I would like to pose a question to the geneticists and mathematicians
subscribing to this forum.

Is it possible that a fallacy was introduced initially into the thinking
regarding Y-chromosome marker mutation rates based in part on the fact that
marker mutations can move back and forth (+1 or -1) over time and was not
properly taken into consideration by geneticists also using "limited"
database sizes??



Dr. David K. Faux, P.O. Box 192, Seal Beach, CA, 90740, USA



www.davidkfaux.org



This thread: