GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-09 > 1127038019


From: "John McEwan" <>
Subject: RE: [DNA] Scottish or Irish or ???
Date: Sun, 18 Sep 2005 22:06:59 +1200
In-Reply-To: <6.2.3.4.0.20050918002403.02924b58@mail.blairgenealogy.com>


Dear John (A Blair)

I will have a stab at this. In part because Blair is one of the surnames
I am interested in (Putatively Britonic Celt origin, highest density in
Renfrewshire for the 1881 census, where it made up 0.35% of surnames.
There were 7778 in total in Britain with Blair surname).

The other reason is quite a few people are asking this question on the
list, and this is basically a "how to" guide.

I went to your site and copied all 14 of the 37 STR profiles and added
them to the modals for R1b, "Irish" R1b modal, and "Scots" R1b modal
using values from
http://www.geocities.com/mcewanjc/phase2.htm corrected for subtraction
of 389i from 389ii values (that is I added 389i to 389ii to create the
normal FTDNA reporting convention).

To do this I used Excel.

I then went to
http://www.mymcgee.com/tools/yutility.html and entered the values and
requested a standard report.

The results show that the Blair individuals differ from the R1b modal by
11-15 markers, "Irish" R1b from 12-17 markers and "Scots" R1b by 13-17
markers but differ from each other by 0-8 markers.

Please note that "Irish" and "Scots" R1b clusters only represent only a
small proportion of the total R1b in these respective countries.

I then went and looked for them in the 37 STR phylograms I have produced
see
http://www.geocities.com/mcewanjc/p3analysis.htm
I found the all Blairs in located within the larger R1b haplogroup, but
most were in a small group see
http://www.geocities.com/mcewanjc/37strallhapr1bfour.pdf (type in Blair
in the search box and after searching magnify the page up about 800%).
The larger group they are tentatively part of is called R1bSTR38 and
this is not either the "Scots" or "Irish" cluster. The Scots cluster is
visible further down the pdf. There are really not a lot of close
matches with the Blairs within the R1bSTR38 cluster (typically 9-11
differences) with the exception of one individual called Strong (Ysearch
number QQUFE) who varies at between 5 and 9 markers. Using Doug
McDonalds mutation rates this individual may have had a common ancestor
with your Blairs ~500-1000 years ago although that is just a best guess.

Now at this point I should make a disclaimer as well. The phylograms are
usually pretty good at clustering closely related people together, and
are also pretty reasonable in sorting people out into different
haplogroups. However, don't view clusters like R1bSTR38 as defined in
stone. Drifting wisps of smoke would be a more apt description. Take a
different random sample of from Ysearch and individuals will move
between clusters. Typically those like yours are at risk, because they
are only poorly related to any other members in the selected group.

So in summary, the Blairs are unlikely to be in the "Scots" or "Irish"
R1b clusters and appear to not be that closely related with other
individuals that have different surnames in the FTDNA database with 37
markers. The exception is QQUFE.

Hope this helps (I will also send you the results as an attached file
directly)

Cheers

John McEwan


This thread: