Archiver > GENEALOGY-DNA > 2006-03 > 1141589647

From: "Alfred A. Aburto Jr." <>
Subject: Re: [DNA] Is the CMH J1?
Date: Sun, 05 Mar 2006 12:16:56 -0800
References: <>
In-Reply-To: <>

> Bonnie Schrack wrote:

> Sorry I hadn't seen Al's posting yet, when I wrote earlier.
>> ________CMH J1 J2
>> DYS393 = 12 12 12
>> DYS390 = 23 23 23
>> DYS19 = 14 14 14
>> DYS391 = 10 10 10
>> DYS385a = 13 -- --
>> DYS385b = 15 -- --
>> DYS426 = 11 -- --
>> DYS388 = 16 16 15
>> DYS439 = 12 11 11
>> DYS389I = 13 13 13
>> DYS392 = 11 11 11
>> DYS389II = 30 30 29 (these are not right because I should have done
>> 389II-389I)
>> (couldn't fix 389II right now because my computer crashed, CPU fan
>> went belly-up)
>> (will fix tomorrow)
> Nice job, Al! You've found the modal J1 and J2 haplotypes all right,
> the same ones that show up in my data, which is not all SNP-tested of
> course. DYS426 is of course 11 in all of J. I haven't had a chance
> (yet) to calculate DYS385 as a connected single marker, but DYS385a
> has a modal of 13 in J1, J2a1*, and J2a1k. The modal for DYS385b is
> 18 in J1, 15 in J2a1* (which could actually change if I enter more
> data), and 16 in J2a1k. But people should not dream of using that
> marker to distinguish clades, since it's highly variable.

My computer, offline now, with all my data for the above, will take
longer to fix, but when I get it fixed, I'll send you the locus
distributions for J1 and J2. I did J too. I'll send that too. The
results are interesting to me. DYS388 for example surprised me because
it has such a large spread over alleles. I think that indicates it might
be a fast mutator. Others have very little spread, indicating slow
mutators. DYS385a looks very steady, but DYS385b shows more spread. Same
with DYS389I and DYS389I.

>Bonnie wrote:
>You're stating in another way what I said: the 12-marker CMH, we're
told, is 12, while the J1 modal is 11. J1 also normally has slightly
higher >DYS388 values than J2, with a modal value of 16, as seen in the
CMH. Anyway, the point is that the CMH haplotypes are close to the J1
>modal, and this can also be seen at many other markers, when you look
at 25- and 37-marker haplotypes which have the CMH in their first 12
>markers. The correlations in the markers beyond those first 12, make
it clear that they belong to J1 and not to J2.

Humm, interesting. I just, during this discussion, learned of the 12
marker set for the CMH. I didn't know you already knew of it Bonnie.
I think it would be best to ignore the 12 marker set until something
more official is published. At least ignore it for any type of further
research work.

> Now on to the next:
>> I did't realize DYS413 was a UEP, a SNP! So then there appears to be a
>> challenge here that the CMH is specific to J1?
> Not really. DYS413 can certainly be considered a UEP, but that
> doesn't make it a SNP -- Dienekes is saying that the deletion at
> DYS413 is another kind of Unique Event Polymorphism, although it's not
> a Single Nucleotide Polymorphism -- this is correct.

Well, Malaspina's paper is indicating "yes, indeed CMH is part of
J2a1", but still from what I can tell their case looks weak, not a
strong argument at all. But they are making a strong statement. I mean
they were saying these seemingly J2 Italian Cohanim had the DYS413<18
constraint. However, that said, it doesn't seem right to ignore
published technical material (Malaspina's paper) over that which has not
been published or peer group reviewed (FTDNA). I wouldn't take FTDNA's
12 marker set as scientific evidence until it appears in the open peer
group reviewed literature. They are _not_ following standard scientific
practice and methods.

>> Yes, the 6 Y-STR CMH is found in both J1 and J2. Also, if one uses
>> FTDNA's (unpublished!) 12 Y-STR CMH then it appears that J1 is indicated
>> instead of J2. However, as Sasson said, apparently it is not known if
>> the real Cohanim, and therefore the source of the CMH, are J1 or J2. I
>> think it is certain that SNP testing has been done on Cohanim, but no
>> one is talking. It appears to be a very sensitive issue, so I can
>> understand the secrecy or perhaps hesitancy to publish such knowledge...
>> A 12 or more Y-STR CMH would clarify things greatly. However, FTDNA's 12
>> Y-STR CMH must be treated as unsubstantiated until it is published in a
>> peer group reviewed open literature scientific journal. The whole issue
>> surrounding the CMH remains, well, fuzzy ...

> Well, for now, anyway, I choose to accept FTDNA's 12-marker CMH as
> basically accurate; I think they have access to information on real
> Cohanim families, even if they don't feel that they're yet in a
> position to publish the substantiating evidence. It could well be
> that this is a preference of people or communities who are directly
> involved? Anyway, I don't think the issue is really so fuzzy or
> unclear as it's being seen in this discussion.

But be aware that things might change on you, because FTDNA is saying
things and doing things outside of normal scientific preview and even
outside an open forum (like here).

> David wrote:
>> YAP identifies the DE branch of the tree. It doesn't apply to J.
> It was just one of the SNPs they tested for, in those early days, to
> place their samples in a haplogroup.
>> There was a recent study which reported finding the CMH pattern in
>> BOTH J1
>> and J2. It was mentioned on the list about six or eight weeks ago, if my
>> internal timekeeper is working this morning. This outcome would seem to
>> require reinterpretation of the Cohanim data, or even a search for
>> additional data that might help draw finer CMH distinctions than we have
>> right now.
> David, any number of studies can easily show that a 6-marker CMH
> haplotype can be found all over the place. The point is that the
> 6-marker CMH is not very meaningful, precisely for that reason. The
> fact that it's found in J2 as well as J1, doesn't require
> reinterpretation of the Cohanim data (unless I mistake your meaning),
> it just means that we need more markers to have a meaningful CMH. And
> now, we do, albeit unofficially. Hey, how "official"are any of John
> McEwan's STR clusters? The fact that they are unpublished doesn't
> detract an iota from their usefulness. We can all think of more
> examples of useful modal haplotypes discovered by genetic
> genealogists, which have not been published. Why should this be held
> to so much more exacting a standard?

I'd say that by FTDNA's methods, it is very unofficial. John McEwan's
STR clusters are done in a much more open environment I'll bet, not just
announced as FTDNA did ...


This thread: