GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2009-12 > 1260241483


From: (John Chandler)
Subject: Re: [DNA] R-U152 and R-L21 on the European Continent
Date: Mon, 7 Dec 2009 22:04:43 -0500
References: <59915.56775.qm@web81103.mail.mud.yahoo.com><06F79C98-398C-4200-8D9C-5BA976ABD233@vizachero.com><ea3bd9560912061730y581da805v54dec4a73733efee@mail.gmail.com><006701ca76ef$d0ab50b0$7201f210$@org><ea3bd9560912071008k310d198dre58efe3f4d2676cc@mail.gmail.com>
In-Reply-To: <ea3bd9560912071008k310d198dre58efe3f4d2676cc@mail.gmail.com>(message from David Faux on Mon, 7 Dec 2009 10:08:29 -0800)


David wrote:
> Perhaps it is necessary to turn to the chemistry of the system to understand
> why it is a weak assumption to take a snapshot of what is evident from today
> (e.g., via father - son observation of mutation rates), and say that it must
> have been so in the past. The more I learn about the process of
> methylation, and the role of histones in gene expression and other genetic
> mechanisms, the more it becomes apparent that environmental factors (which
> of course have not been constant over the eons) can impact on what occurs at
> the genomic level.

Basically, you are proposing to throw out the "baby" and keep only the
"bath water". Any and every conceivable calibration method for age
calculations must be anchored by measurement of the known -- i.e., the
present day. If you give up the assumption that the present day is
representative of the past, you have no basis for any calculation at
all. Ever. The fact is that we already have a window into the past
provided by deep-rooted pedigrees, and the mutation rates calculated
from pedigrees are *not* subject to a Zhivotovsky-like fudge factor.
Indeed, the Somerled pedigree covers roughly the same time depth as
the case studies used by Z et al. Some time ago, Doug reported a
curious fact about the Clan Donald results -- counting actual
mutations corresponding to actually documented meioses produced an
average mutation rate consistent with other pedigree-based and
father-son-based estimates, *but* taking the variance of all McDonalds
and kin without regard to how they are actually related came up with a
significantly lower estimate. What this tells us is that the past is
pretty much like the present, as far as we can tell, but that the
details of population history are crucial in intraclade age
calculations. The fact that the details are unknown in most cases
means that intraclade ages simply cannot be calculated in general. At
best, such ages can be bounded.

John Chandler


This thread: