DNA-R1B1C7-L Archives

Archiver > DNA-R1B1C7 > 2011-04 > 1303947825


From: Allene Goforth <>
Subject: Re: [R-M222] Selected M222 surnames
Date: Wed, 27 Apr 2011 16:43:45 -0700
References: <2dcd2.1b63c2d6.3ae9de16@aol.com>
In-Reply-To: <2dcd2.1b63c2d6.3ae9de16@aol.com>


Hi John,

Yes, MacAdam is a Scottish surname. Although the five lines were living
in the area near Arisaig in the Highlands (Glenuig and
Arieniskill/Arienskill) at the time they emigrated to what is now Prince
Edward Island in 1790, they all have some matches to the Milligans and
Griersons in the Galloway area. In addition, they have eight 11/12
matches with McAdam/McAdams. I haven't tried to find out anything about
those people yet. Been too busy trying to sort the aforementioned five.
It'll be interesting to see their 111 marker results. I'm glad Bill
Howard is taking an interest in my little cluster.

Allene

On 4/27/2011 2:01 PM, wrote:
> Bill Howard (who came up with the M222 phylogenetic charts posted
> recently) has taken some of the major surname clusters from the M222 project for
> closer analysis. The surnames he chose all have multiple entries, most five
> or more.
>
> I wonder if anyone on this list has any information about some of the
> surname clusters in the project?
>
> Ewing (22) - As far as I know this is a Scottish surname. The DNA
> participants in the project trace themselves to a Ewing family living in the
> vicinity of loch Lomond in lowland Scotland. Beyond that nothing is known for
> sure.
>
> Dunbar (8) - I know this surname is Scottish but the M222 Dunbars
> represent just a small fraction of the Dunbars tested in the databases. The only
> one with any definite ancestor location says Tyrone.
>
> Doherty (18) - Irish, Cenel Conaill, ie, a descendant of Nial.
>
> McGonigle, Megonnigal (9) - Irish? McGonigle is a common Donegal surname.
> They appear as herenaghs of churchlands in two different parishes in
> Donegal.
>
> Howle (7) - The Howles in the project list their earliest known ancestor
> as Epaphroditus Howle, b.c. 1685, New Kent Co., VA England. The surname
> profiler shows the name is peculiar to southern England. There is some
> speculation on personal web sites the name may be derived from the Welsh Ap
> Howell. Anybody have a clue about this one?
>
> McAdam (5) - Is this a Scottish surname? I saw some speculation recently
> that this may be a Galloway surname,
>
> McCord (8). Scottish? I think this has also been discussed as a Galloway
> surname.
>
> Cowan (17) - Scottish? Some of the Cowans in the project list Scotland as
> place of origin. Others list Ireland.
>
> Bill has a phylogenetic chart composed of these surnames from the project.
> It shows the Ewings splitting off the M222 tree in about 85 BC or prior
> to the time of Nial. In this grouping they occupy one segment of the M222
> tree by themselves.
>
> The charts then show all of the other surnames with a TRMCA of about 470
> AD. which is close to the time of Nial (c. 450 AD). They fall into two
> broad groups.
>
> 1. Dohertys, McMonigles, Dunbars
>
> 2. Howles, McAdams, McCords, Cowans.
>
> The Doherty-McMonigle cluster makes sense. Both were probably Cenel
> Conaill from Donegal, descendants of Nial. I don't know why the Dunbars cluster
> in that group.
>
> The Howles and Davidsons (I forgot to mention them) have their own
> separate cluster within the second group. McAdamds, McCords and Cowans are
> cluster together with a common ancestor at about 940 AD.
>
> In the larger chart posted on the M222 web site the Ewings are clustered
> with several other surnames, some of which don't make much sense to me.
>
> O'Flanagan, McLaughlin, Doherty, Vanover (this name is really Mullins),
> Mitchell, Nicholson, Alexander.
>
> On the larger chart there is a complete segment composed primarily of
> Dohertys and McLaughlins with a few other surnames mixed in. It begins with
> Fleury (kit 105896) and runs through McLaughlin (kit 90801). The McGonigles
> are also clustered in this group. This make perfect sense since these
> three Donegal surnames were said to be descended from Nial.
>
> On the larger chart Cowans and McCords are mostly clustered in one
> segment. But in this chart the Dunbars are in a cluster segment far from the
> McLaughlins.
>
> I don't really know what to make of these charts yet.
>
> Bill's charts are not based on genetic distance as are most similar charts
> I've seen (Phyllip). It compares each sample to every other one in the
> database assigns them a similarity value, then groups the samples accordingly.
>
> A member of the list describes the process better than I can:
>
> Let me try to explain how I understand this research and the associated
> tree. I hope it is helpful.
>
> A basic statistical method is to calculate a value called correlation. This
> is used to see if there as a link between measured variables. For example,
> to see if there is a correlation between a person's height and how many
> years they live. I'll skip the mathematics, here, but running this
> calculation measures how strong the relationship between the variables is
> (height and longevity in the above example). This function is built into
> Excel and is called "correl". Bill Howard used this function on the M222
> STR
> data to calculate the correlation between every possible pair of
> individuals.
>
> "Correl" returns a value between -1 and 1, where -1 is no correlation, and
> 1
> is correlated. For two individuals with identical markers, "correl" will
> equal 1. Many of the "correl" values that are returned on the M222 STR data
> are very close to 1. Such as 0.99548. Bill Howard then performed some basic
> math manipulation to turn that number into something more human friendly,
> such as 45.39. This last number is what he calls the RCC (revised
> correlation coefficient). His next step was to use his method on
> established
> pedigrees to determine a relationship between RCC and time, and he came up
> with about 43 years = 1 RCC. So, if two M222 individuals have a "correl" of
> 0.998, the RCC = 20.04, and the presumed time to a common ancestor is 20 *
> 43 = 860 years (approximately).
>
> At this point I am speculating, but I interpret the chart thusly:
> individual
> M222 members are clustered with others that they have a low RCC to. Recall
> that the RCC is not a unique value for each person, but that there is an
> RCC
> between each possible pair in the M222 dataset. Clusters with low RCC
> relative to one another should, in theory, represent individuals that are
> more closely related. Bill Howard used a program called Mathematica to sort
> all of this out, and the phylogenetic tree as presented is the result.
>
> One thing to note is that the results do not use a genetic distance, or a
> modal. "Correl" is math: means, squares, and multiplication. For a small
> enough marker set, it is possible to decrease the genetic distance between
> two individuals but make the "correl" worse, the opposite of the expected
> effect. This may be the source of some unexpected relationships on the tree
> or, then again, maybe not (as he uses 37 markers).
>
> Bill did compare his results using 12, 25, 37, and 67 marker sets. He found
> that there wasn't much difference in the end, and used the 37 since there
> were more samples (although going under 37 increased the sample size, it
> increased the uncertainty in the values).
>
> PS: Two random M222 individuals with 67 markers (Kits 91704 and 150141)
> have a calculated "correl" of 0.99870 and a corresponding RCC of 13.06. The
> genetic distance is 7. If you "fix" three of the mutations (chosen at
> random), the GD drops to 4, and the "correl" becomes 0.99926 for an RCC of
> 7.43.
>
>
> The process isn't perfect. You'll see some Dohertys scattered around in
> random clusters with other surnames in his large chart. These are probably
> outlyers. Some of our McLaughlin samples are also scattered about in odd
> spots on the charts. But the core group of each does appear in uniform
> cluster segments.
>
> For those surnames in the charts with only one or two samples I think
> caution is needed in interpreting their position on the charts.
>
>
>
> John
> R1b1c7 Research and Links:
>
> http://clanmaclochlainn.com/R1b1c7/
> -------------------------------
> To unsubscribe from the list, please send an email to with the word 'unsubscribe' without the quotes in the subject and the body of the message

This thread: