GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2007-02 > 1171559956


From: "Terry Barton" <>
Subject: Re: [DNA] How do you do an estimated Common ancestor
Date: Thu, 15 Feb 2007 12:19:16 -0500
In-Reply-To: <mailman.20814.1171507376.14078.genealogy-dna@rootsweb.com>


Orin, I agree with you. There can be no precision in a MRCA calculation.
Stating it as I think - Individuals do not have an average number of
mutations - and therefore, a calculation of distance to MRCA must be a range
of possibilities.

I offer my own results as a fairly extreme example:

* My uncle Bob and I are 41/43 - 2 mutations of difference in 3 birth
events. By triangulation and extensive testing, we know that one of the
mutations was introduced with my dad and the second with me.

* My uncle Bob and Richard are 43/43, but cannot share a common ancestor in
less than 11 generations or an common ancestor born any later than the
1620s - based on paper trails.

* A cursory look at results - without looking at paper trails - would
presume that my uncle Bob and Richard (43/43) were closely related and that
Bob and I (41/43) were not nearly as close.

* If any reader cares to do the MRCA calculation (pick your favorite
calculator and favorite mutation rate - it won't matter much)

** you'll find that that there is a very high probability of a shared
ancestry for Bob and Richard at 11 or more generations.

** you'll find that the calculated probability of my uncle Bob and I
sharing a common ancestor in 1.5 generations is very low.

** and you'll find that the 50% probabilty of a shared ancestor for 43/43
and 41/43 is just a number - wrong in both cases - if you are trying to use
it as a precise answer

A further caution for List Readers wanting to obtain precision from their
comparisons - which markers are tested can also affect the comparison:

*** when my uncle Bob and Richard are compared on FTDNA's 37 markers - the
"outcome" is different - as they are different on CDYb - giving them a 36/37
match. (instead of 43/43)

*** Additionally - my mutation occurred on a marker not included in the
FTDNA 37 - so my uncle Bob and I are also 36/37. (instead of 41/43)

*** Slightly different markers and "different" outcomes - as now we seem to
be the "same" closeness - again wrong.

Three men - two sets of compared markers - differences in the matching
levels. Needless to say - I don't pay a lot of attention to pecisely how
many mutations separate two men - I am much more interested in which markers
show a mutation and how each man compares to the ancestral haplotype.

As a further trivia - my dad and I are 11/12, which extends out with more
testing to a projected 112/113. You'll get a different 50% probability with
each different comparison for our MRCA at 12, 25, 26, 37, 43, 48, 67, 76 or
113 markers.

More trivia - My son and I are 113/113 (I have been trying to find a
mutation for my son - no luck yet). But - despite any MRCA calculation - I
am as closely related to my father as to my son.

Best regards, Terry Barton


Message: 2
Date: Wed, 14 Feb 2007 14:07:12 -0800
From: OrinWells <>
Subject: Re: [DNA] How do you do an estimated Common ancestor

I know I am apparently the only contrarian on this point because so many of
the members of this list seem to think there is some magical precise science
and math involved here. There just isn't It is totally random. If anyone
can show me where this has paid off in more than one or two equally random
cases, I might be willing to say I am wrong. But it seems to me to be such
a waste of time.

This thread: