DNA-R1B1C7-L Archives

Archiver > DNA-R1B1C7 > 2011-12 > 1324622821

Subject: Re: [R-M222] Ulster
Date: Fri, 23 Dec 2011 01:47:01 -0500 (EST)

In a message dated 12/22/2011 12:01:37 P.M. Central Standard Time,

Apologies for the delay in replying. With a family of four, Xmas is now a

To explain, in October Paul and Sandy exchanged comments about the
likelihood of a SNP downstream of M222+ and Sandy replied with an
comment, 'that he thought M222+ is highly unlikely to be much more than
50 generations 'old' and that it is therefore possible that any SNP's
downstream of M222 may be classified as 'private' by whoever makes these
decisions'. If the mutation rate of SNP markers is extremely slow
(according to
FTDNA's own webpage, mutations are rare and happen at a rate of
one mutation every few hundred generations), the likelihood of one
occurring downstream in this generation seems even lower if the M222+ is
only 50
generations old.

50 generations at 30/gen. is 1500 years or about 500 AD. which is about
equal to the typical ASD TMRCA estimates for M222. Some might come in a hair
earlier but not by much. I warned Bill numerous times his 1680 BC. date
would be highly controversial. He didn't seem to care - such is his faith
in his own methodology, which I do not share. When the Walk on the Y began
several DNA experts including Nordtvedt warned us M222 was such a young
clade nothing new was likely to be found. I'm sure that was based on the
same mutation rates you mention.

A new SNP downstream of M222 would be nice - but I'm not holding my
breath. Several M222 samples have been analyzed under the Walk on the Y
program including one just a few months ago. Nothing was found. There is a
reason why so many in the DNA universe dream of new SNPs. The fact is they
can't tell anything from STRS alone. That's a problem in every large
haplogroup project, including M222. The Beatty/Byrnes project is composed of
mostly Irish and Scots with some English. The Irish have been linked with the
Lagin chieftains of Leinster. But no one has any idea of why there are
matches in Scotland such as the Beattys. No one has the slightest idea
where the DNA originated, whether in Ireland, Scotland, England or Gaul. They
did discover a new SNP for the clade through Walk on the Y (L-159-2). But
prior to that they were just L-21. The Irish Type III project is also
tied in with an Irish tribal line (the Dal Cais). They also found a new SNP
in WTY - they also have completely unexplained matches in Scotland and
elsewhere in the British Isles. No one has a clue where it originated.

I'm mentioning these other projects to demonstrate that our lack of any
real knowledge about the origins of M222, frustrating as it may be to
people, is not unique. In fact it's par for the course.

<I really do hope one is found and if not, those on this forum who are
following the DF23+ M222- marker trail, find another way round to identify
possible origins of M222 SNP.

I don't see much there yet. One interesting sample is:

N92711 Le Provost (Avranches Manche ca1580-1640)

Otherwise there are some in Ireland, a few in England; one in Scotland.

<Arguably you could also include the Ewing and O'Docherty clusters and of
course your own surname. As you know there already exists a proven paper
trail for the chief family of O'Docherty from Spain to Co. Cavan in the
The only copy of the DNA test I have is a 37 marker test (do you know if
it has been extended?). This one holds a pedigree that follows the chief
line from Co. Cavan back to the Cenel Conaill in Co. Donegal. I must admit,

it has been awhile since I looked that this pedigree on my own website, so
there may be new information on it that I am not aware of. The O'Docherty
lineage can be reached on the following link and is on the right column on
the bottom section:

That is a pretty reliable pedigree although not fully proven in the
genealogical sense. There are a few undocumented generations in Spain but
overall I see no reason to doubt it's validity. The real undocumented portion
of the pedigree is found in the generations from Nial to about 1200 AD.
But as I recall it does match up well in terms of generations to the O
Domhnall pedigree. The MacLochlainn/O'Neill pedigree is much better documented
but it too has problems. I show about 34 generations back to Nial from
Brian oge McLaughlin, alive in 1601, the chief of sept. That compares to
about 35 generations back to Nial from Sir Cahir O'Doherty of about the same
date. Add another 12 generations or so to the present day and you get 46
generations for McLaughlins and 47 for Dohertys. They're both about the
same. I'm not sure what any of this proves except that the Irish pedigrees
aren't glaringly off in generations from Nial.

<Looking at this pedigree traced from Cathaoir O'Dogherty (I am aware
may be gaps earlier), Sandy's estimate of 50 generations isn't far of the
mark. As you know in offline discussions, the Milligan/Milliken-Grierson
DNA suggests the common ancestor here, is almost certainly over 30
generations ago and is localized in Mid Nithsdale prior to AD 1100. If the
of 50 generations (allowing for marginal errors), 20 generations back to a
common ancestor from Nithsdale isn't that long by any estimate. Do we have
two identifiable subgroups with the M222 haplogroup? I think the answer is

I'm not sure which two subgroups within M222 you're referring to here.
The Dohertys in Ireland and the Nithsdale cluster in Scotland? We can use
the Dohertys rather than the Cenel Eoghain McLaughlins. Makes no
difference to me. The common ancestor of the Dohertys was Aindilis O Dochartaigh,
d. 1292. That's according to the O'Clery genealogies. The surname is older
than that - how old we do not know. There aren't any records The surname
first appears in the annals in 1180 with the death of an Aindilis
O'Doherty. The Dohertys rose to prominence at the same time as the O'Donnells.
The Dohertys have a well defined family modal which the DNA sample you refer
to fits perfectly (a known descendant of Sir Cahir O"Doherty, slain in
1608. They are identifiable in DNA through the family modal but does this
make them a sub-group of M222? I suppose you could say it does but then every
family cluster with a modal would also be a sub-group. By that definition
the Ewings and McLaughlins would also be sub-groups. The problem I have
always had (and still do) with DNA is I can find nothing to link these
diverse sub-groups together as related except for the M222 modal. I see nothing
for example that tells me the Dohertys and McLaughlins are more closely
related than surname groups from Scotland or elsewhere in Ireland. Or that
the McLaughlins and O Cathains are more closely related to each other than
to Dohertys.

<Bill Howard in his recent article on 'A Dated Phylogentic Tree of M222'
drew an interesting conclusion by writing, a 'lack of attention to possible

concentrations of the M222 SNP in Scotland and a lack of definition to how
the Ui Neill descendants can be differentiated from carriers of M222 SNP,
have hindered research into its place of origin as well as finding
ties to Niall and his Ui Neill descendants as a separate entity from the
descendants of the M222 progenitor'. Not everyone might necessarily agree
with his estimated age of the M222 SNP, but I do think his concluding
comments are absolute right.

That's what Bill Howard is referring to when he talks about a lack of
definition among the Ui Neill vs. other M222. At this point I think that's
just a fact of life with STR analysis. I had some hopes for the new 101
marker set but so far don't see anything useful myself. Maybe they just haven't
found the right markers to test yet?

Another approach to the dilemma of identifying possible sub-clades
within M222 is Bill's method, or if you don't like that, then the traditional
genetic distance clustering by programs such as Kitsche and Phyllip which
amount to the same thing. The problem with that is no one except Bill knows
how to employ his own method and the Phylip software appears to be
limited. I tried to run the entire M222 project through Kitsche and Phylip but
the software hung up after 12 hours. It appears the program can handle only
so many samples at a certain resolution. There might be some software
program out there that can handle large datasets at high resolution but I do
not know what it might be. Almost every project out there uses both Fluxus
and some kind of Phylip-style clustering in their analysis including the
L-21 project.


This thread: