GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-09 > 1128012599


From:
Subject: Re: [DNA] STR mutation rates
Date: Thu, 29 Sep 2005 12:49:59 EDT


In a message dated 09/28/05 1:21:38 PM Pacific Daylight Time,
writes:

> Note the peculiarity of the log data: the average for 1-25 is higher
> than either 1-12 or 13-25. This suggests a bias in the decisions to
> upgrade from 12 to 25 -- those with observed mutations in a 12-marker
> test were significantly more likely to upgrade than those with a 12/12
> match. In particular, the average rate for the 12-marker tests of
> those who upgraded is 0.00267, while the average for those who did
> not upgrade is 0.00175. This effect would be even stronger if we
> could count the 11/12 near-matches that exploded upon upgrade and
> were therefore eliminated from the statistics.

That's an interesting observation about the data in Charles Kerchner's log. I
was trying to think through whether this effect might have an impact on
FTDNA's mutation rate study, especially for the 3rd panel (which I am assuming has
a smaller number of samples). I didn't come to any conclusion, though.

As for the overall rate in the log, I haven't uploaded my data, mainly
because I have a rather small number of transmission events in my known pedigrees.
However, there are no mutations, and if others are behaving like me, that would
bias the log toward a higher rate.

Ann Turner

This thread: