DNA-R1B1C7-L Archives

Archiver > DNA-R1B1C7 > 2009-12 > 1260545840


From: Moore <>
Subject: Re: [R-M222] Coyne and Golden project families
Date: Fri, 11 Dec 2009 07:37:20 -0800 (PST)
In-Reply-To: <000601ca7a65$27e4c4b0$77ae4e10$@com>


He is not the only one disappointed by your diatribe and rude comments. You did not seem to use very tactful language, and I hope you don't rely on your interpersonal skills to make your living. I have over 25 years in the business of serving people in need and you were out of line! J Moore MD

--- On Fri, 12/11/09, Sandy Paterson <> wrote:

From: Sandy Paterson <>
Subject: Re: [R-M222] Coyne and Golden project families
To:
Date: Friday, December 11, 2009, 8:23 AM

I am well aware of the fact that the subject of DNA genetics is still in its
infancy. I am also well aware that some people on the list are new to the
subject and are on the list in the hope that they may learn a little and
find things of interest.

However, I find your smugly pedantic utterances about probability theory
quite irritating. Perhaps if you dropped the pretence that you actually know
something about the subject we'd be able to make a little progress.

@     
60 generations without a mutation?  Hardly likely.
@

Quite. This is typical of the way you reason. I have never claimed a
likelihood of zero mutations. In fact I said exactly the opposite.

I quote from a posting on Tuesday, 8 December, 3 days ago :


"A modal suggests nothing. It is merely a logical guess at the haplotypes of
the 'founder' of the haplogroup and the thinking goes that all people who
are M222+ arrived at their own haplotypes by starting off at the modal and
mutating to their current haplotypes. Being close to the modal does not in
any way suggest a closer than average relationship to Niall. It means
nothing.

Take Mackenzie 26917 as an example. It is easy to fall into the trap of
believing that he started off with the M222+ modal, had no mutations, and so
ended up on modal.
In fact that is extremely unlikely. The probable number of mutations over 67
markers in one generation is about 0.2243. That is the sum of the 67
mutation rates currently used by most genetic genealogists in simulation
work, and comes from the work done by a chap called John Chandler.

If we take the 'age' of M222+ to be 60 generations, the probability of
having zero mutations over 67 markers in 60 generations is

(1 - .2243)^60 = .0000002407839

Virtually zero."




@
You don't know the haplotype of the original male, yet you ASSUME that the
modal you have calculated
from a relatively tiny sample of R-M222 is the original male's haplotype,
and you proclaim (without
evidence) that those who match this small-sample derived "modal" 67-67 have
preserved or reverted to
this hypothetical haplotype and defied the laws of probability across
potentially hundreds of lines
and 60 generations? 
@

If you had even the vaguest idea about mathematical modelling, you'd realise
that a modal observed in a current population is the best estimate of the
founder's haplotype. It should also be obvious that the probability of
changes of modal values over 60 mutations is small. So the majority of the
founder's marker values should be retained as the group modal.

I've done a few simulations that confirm this. I need to do a few more and
will post links so that any interested list members can examine the
simulated haplotypes once I've completed the exercise. So far, the founder
haplotypes have been retained at the following rates (I used the M222+ modal
as the starting modal):

65/67
64/67
63/67
65/67
65/67
67/67
65/67
65/67

As could have been expected, the changes tend to occur with a higher
frequency amongst faster mutating markers. Of the 17 changes, 5 occurred in
CDYa,b.

Defied the laws of probability? More jargon borrowed from a subject that you
know very little about.




@
Even if you do accept this (which is not at all parsimonious), it is all
dependent upon your
hypothetical modal being correct (exactly correct).  If it isn't, and the
actual ancestral haplotype
is that of one of your "significant" multi GD off-modal matches, then it is
the latter who are no
longer of determinable relatedness, and the former who are now regarded as
significantly related.
And remember, as the sample size and diversity increase, the calculated
modal changes!
@

My goodness. More gobbledygook, sprinkled with a red herring or two.

If you want to investigate possible kinship, you need to look for rare
characteristics. If two or more people in a population share one or more
rare characteristics there is a chance that they may have inherited them
from a common ancestor. The rare characteristics (off modal matches in this
case) may of course have come about by inheritance from two different
ancestors (so-called parallel mutations). But the greater the number of
unusual characteristics, the smaller the probability of the characteristics
having come about by chance.

The sharing of characteristics that are common to the M222+ population group
(such as modal matches) are of no help in identifying possible kinship,
simply because they are common.

And this has absolutely nothing to do with whether or not the modal
replicates the founder's haplotype. I has to with rare characteristics.

@
Your "logic" regarding how we cannot draw conclusions from *lack* of
evidence should be equally
applied to your analysis of evidence based upon unproven assumptions.  There
is no evidence in
unproven assumptions or estimates.
@

Another classic.

On reflection, I think 'gobbledygook' is more accurate than 'rubbish',
although it's a tough call.


Sandy





-----Original Message-----
From:
[mailto:] On Behalf Of Gregory Brown
Sent: 11 December 2009 08:09
To:
Subject: [R-M222] Coyne and Golden project families

Sandy,

Now you're just being nasty.  I don't believe that referring to good-faith
comments relating to
probability, genetics and population dynamics as "plain rubbish" is very
constructive or a good
demonstration of your level of self-proclaimed expertise...well, perhaps the
latter, but certainly
not the former.

And "goes beyond plain rubbish" is just plain rude.

If you wish to hang your authoritative hat on circular reasoning by using a
living population sample
to erect a modal haplotype, and then use that same sample-derived modal
haplotype to deny the
significance of a 67-67 match, then you may as well stop wasting your time
assuming DNA testing can
determine relatedness at all.  60 generations without a mutation?  Hardly
likely. Show me a group of
people who can prove their common ancestor back 6 generations who don't have
a GD of 1, 2 or even 3
at the 67 marker level...or even the 37 marker level, and you might get my
attention.

You don't know the haplotype of the original male, yet you ASSUME that the
modal you have calculated
from a relatively tiny sample of R-M222 is the original male's haplotype,
and you proclaim (without
evidence) that those who match this small-sample derived "modal" 67-67 have
preserved or reverted to
this hypothetical haplotype and defied the laws of probability across
potentially hundreds of lines
and 60 generations? 
Even if you do accept this (which is not at all parsimonious), it is all
dependent upon your
hypothetical modal being correct (exactly correct).  If it isn't, and the
actual ancestral haplotype
is that of one of your "significant" multi GD off-modal matches, then it is
the latter who are no
longer of determinable relatedness, and the former who are now regarded as
significantly related.
And remember, as the sample size and diversity increase, the calculated
modal changes!

Occam's razor is sure going to get dull on that one.

Your "logic" regarding how we cannot draw conclusions from *lack* of
evidence should be equally
applied to your analysis of evidence based upon unproven assumptions.  There
is no evidence in
unproven assumptions or estimates.

> I give you my authoritative opinion that you are way off
> base. If you want a
> second opinion, contact me tomorrow and I'll give you a
> second opinion with
> the greatest of pleasure.
>
>
> Sandy

Thanks, Sandy, I'll pass.  You don't owe me a second opinion, regardless of
how much pleasure it
would give you.   But, as a very active member of this list, you DO owe the
members a bit more
substantial explanation for what you are espousing. Maybe a reference or two
to a papers which can
explain why a 67-67 modal match between two people is insignificant in
determining relatedness?
Now, THAT would be appreciated.

Greg



 



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: