Archiver > GENEALOGY-DNA > 2003-11 > 1067874193

From: Mike Humphrey <>
Subject: RE: [DNA] Marker Mutation Rates [and MRCA]
Date: Mon, 3 Nov 2003 07:43:17 -0800 (PST)
In-Reply-To: <>

David, Jerry ...

I'm very happy to see this recent MCRA discussion. During the past 7 months
since I started reading this list, I've been struck by the wide variance in
observed mutation rates in several surname projects. E.G., Doug Mumma's
excellent study in which he saw rates faster than .002.

I agree completely with David's statement that the concept of mutation rates
(& MCRA) need a lot more refinement before seriously relying on them for
genealogical research. But I must admit that I enjoy the "statistical game",
comtemplating the theoretical possibilities. As many have said, the DNA data
is valuable as a tool to help support a hypothesis based on paper trail

In the future I will crank up a surname project with FTDNA to attempt to solve
some puzzles (e.g., who were Michael Humphrey's ancestors - Lyme Regis, Dorset,
1590- ?). Meanwhile, I'm enjoying my 6 exact 12-marker matches & 23/25-marker
matches with Suttons & Lisles.

(Like some others, I tried the OA Y-search yesterday. Had 2 matches:
both from the U.S., with deeper ancestry in "Prussia" & Switzerland).

Mike ...

--- David Faux <> wrote:
> Hello Jerry and all:
> Yes, it was after reading your earlier post that I became more wary of mutation rate
> calculators and the like. None of these can say anything useful about individual
> cases. The best that can be offered is a probability statement - which has its place
> in population studies - but most on this list are genealogists.

> . . .

> I am also beginning to wonder whether some Y chromosomes (i.e., families) are likely to
> experience more mutations than others and if so, then any formulae may potentially
> apply to some families, but not to others. Irrespective, the fact is that a mutation
> could supposedly occur at any marker in any generation. It is even possible that in my
> family (for example) a mutation has occurred three times at a particular marker in the
> last 2000 years. What we need is a subset of markers known (from empirical evidence)
> to never mutate (for example there were no YCAIIa/b mutations in the Kayser study).
> If, then, my score on this marker differed from that of another fellow then there is
> virtually no chance we are at all related in the Y line. This would stop a lot of
> gnashing of teeth over whether a participant in a surname study is or is not related.
> It would also mean that my score and that of my ancestor 2000 years ago would be the
> same, and that if today this score is only fou!
> nd among
> (for example) the Saami, it would link me to this group (although a "die out" of those
> with this hypothetical score in other populations within the last 2000 years is
> theoretically possible).
> . . .

> As it stands, neither mutation rates, nor the concept of the MRCA have any purpose
> whatsoever in my research. It would take a tremendous amount of refinement to convince
> me to take either seriously.
> As ever, David F., the iconoclast.
> "Jerry L. Ivey" <> wrote:
> Please go back and review my previous posts about my cousin who has a
> two-marker difference from me. It looks as if the change most likely
> came from his grandfather (my g-grandfather) to him in two generations.
> Low probability events can and do happen. They just don't happen very
> often. :>)
> -----Original Message-----
> From: David Faux [mailto:]
> Sent: Sunday, November 02, 2003 12:41 PM
> To:
> Subject: Re: [DNA] Marker Mutation Rates

Mike Humphrey Trombonist & HPC Computer Scientist
650 345-7108

Do you Yahoo!?
Exclusive Video Premiere - Britney Spears

This thread: