GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-03 > 1204478214


From: James Heald <>
Subject: Re: [DNA] ASDs as "unbiased estimators" of TMRCAs
Date: Sun, 02 Mar 2008 17:16:54 +0000
References: <47CAC418.70005@ucl.ac.uk> <47CAD100.5080205@sbcglobal.net><47CAD8FC.6060309@ucl.ac.uk>
In-Reply-To: <47CAD8FC.6060309@ucl.ac.uk>


James Heald wrote:

> Alfred A. Aburto Jr. wrote:
>
>>I do not believe that Thomas et al (1998), "Origins of Old Testament
>>priests", Nature, referred to the ASD as an "unbiased estimator".
>>Please check.
>>Al
>>
>
>
> Ouch! It's true. Good catch.
>
> Thomas et al use the methodology, but don't actually use the phrase
> "unbiased estimator".
>

To add one other comment on the Cohen data, if you put a 5/5 match
through a calculator like
http://dna-project.clan-donald-usa.org/tmrca.htm

it returns a median TMRCA of 35 generations (70 events there-and-back),
ie about 875 years, with a 5-to-95% range of 3 generations to 150
generations (75 years to 3750 years).

So even if (hypothetically) you had managed to trace back all the
lineages as they had coalesced, back to the last two lines standing,
that is the kind of uncertainty that would *still remain* in the number
of years you'd have to go back further, to get to an original MRCA.


Then you'd have the time it takes to get to just two lines standing
(starting from a group that's not all 5/5) on top of that -- so the
actual uncertainty they should have found with a more realistic model
would have been greater still.

-- James.

This thread: