GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-02 > 1266962155


From: "Alister John Marsh" <>
Subject: Re: [DNA] DYS463 and DYS452
Date: Wed, 24 Feb 2010 10:55:55 +1300
References: <FEA8748CE75246AE9C735629D20F103C@HP> <BFECJOAEEPCFBFFLLBGPEEPKGPAA.dna@irishtype3dna.org><5B7B9DC3797F4327AEDF556189B9AAA4@HP><002b01cab477$17ef7f40$47ce7dc0$@org><REME20100223141307@alum.mit.edu>
In-Reply-To: <REME20100223141307@alum.mit.edu>


John,

You are largely correct with your comment...
>>>>>>>
One point that no one seems to have expressed here is that
the "confusion" over DYS452 is not -- repeat NOT -- a problem for the
use of Ysearch as a search engine for DNA matches. The difference
between the residual Ysearch standard and the "high" standard is so
large that anyone who wishes to look for matches involving DYS452
need only make two searches, one with the low value and one with the
high. Those who are near-matches in either search are correct near-
matches.
<<<<<<<

Just a little story, something which happened to me yesterday. I was
alerted to a potentially distantly related Y-DNA line to me of a different
surname. I checked and found a match which at 25 markers looked very close,
apart from DYS464. However when I did a check at more markers, Y-Search
listed the match as 40 mutations different to me. The 40 mutations was made
up of 19 steps at DYS452, 2 steps at DYS463, both calibration issues, plus
another marker where the person had not adjusted for Y-Search standards (I
think 5 steps on one marker, and possibly one step on another marker?), plus
a recLOH event which went through many of the multi copy markers. A couple
of the very fast mutating markers had multi step differences which impacted
quite a bit on the count, but may have been single multi step mutations.

At the end of the recount, in spite of the initial 40 mutation step
difference by Y-Search, this person was not too many "real" mutations
difference to me, close enough to "possibly" be related in the past 1,000
years.

One wonders if some people have concluded that brothers are not related if
they tested at different companies, and these series of calibration errors
and recLOH etc are heaped up together. People less familiar with
genealogical DNA might be less able to spot the meaningless differences.

It will be as we said about 5 years ago...."good when the calibration issues
are cleaned up between different test companies and different databases".

John.



This thread: