Archiver > GENEALOGY-DNA > 2010-02 > 1265861740

From: "Anatole Klyosov" <>
Subject: Re: [DNA] Variance Assessment of R:U106 DYS425Null Cluster
Date: Wed, 10 Feb 2010 23:15:40 -0500
References: <>

>From: "Lancaster-Boon" <>

Dear Andrew,

Sorry for the delay. I think I start to realize why we are going in circles.
We indeed talk past each other. Here is the last example. I asked:

>Can you specify your question?

You responded:

>I do disagree that I never asked any clear question.

(see below). Can you see the pattern?

O.K., let's move ahead.

I have read your message in its entirety, as promised. Honestly, I do not
know what to say. I have a feeling that you have confused literally
everything related to mutation considerations and counting in haplotypes.
However, I know what will follow. You will state that you do disagree with
that, etc., etc., etc.

Let's agree on the following.

1. Please do not pedal that there are too many complications in "family
trees" and mutation counting, and how much we should know on clades, "family
tree structure", etc. JUST DO IT.

2. Please take a close look at the latest concrete example I gave in this
and in the "MRCA Estimates". Please do not tell that those example are not
those you would like to see. Just get their logic. They in fact answer most
of your questions.

3. No, you do not need "a full family tree" in order to count mutations. All
you need in two things: (1) an average number of mutations per marker, and
(2) an assessment of how many separate (major) branches might be on the
tree. For this you do not need "a full tree structure". An example was given
in my brief analysis of John's example with two branches, one had only two

4. >"In order to count mutations I think you need a full "family tree".

Absolutely not. You will never get a "full family tree" for "old" TMRCAs.
You need just to figure out was there one nearest common ancestor for your
dataset, or two or several of them. I have explained how.

5. Actively use the logarithmic approach when possible. You will see how
easy you can calculate TMRCAs. Furthermore, if the logarithmic and the
linear methods give you the same results, it will give you an additional
comfort. You will see right away that you count mutations properly. If
results are different, it is even more important. You will understand that
something is wrong with your system.

6. >"You need knowledge of not only one common ancestor for whole
uncontroversial clades, but also the common ancestor for each sub-branch and
sub-sub-branch, etc."

Absolutely not. If a tree is non-complicated one (meaning, has only one
common ancestor), just one figure, that is an average number of mutations
per marker (or per haplotype) gives you an immediate answer in terms of
TMRCA. This would be a classical first-order reaction. An example: if you
have a radioactive material and know its decay rate constant, you do not
need to know history of its preceding radioactive decay. It is the same - in
terms of dynamics of the process, as the future decay. It goes according to
the first-order reaction. Those reactions are described with only one rate
constant. The same is with an uncomplicated haplotype dataset.

In other words, the most important job there is to determine, is it an
uncomplicated (first-order) system or a complicated one.

7. >"You give your own small example...".

It does not matter, small or big. If you can handle small examples, you can
handle large ones, and vise versa. By the way, examples by John and Charles
were not "mine".

8. >"And we know that true mutation counting requires knowledge of all the
nodes in this descent down to today..."

Absolutely not. See above. That is why I have applied the logarithmic method
to datasets analysis. Take a look at the Charles example.

9. >"...others find it very hard to be confident about this".

I do not by "arguments" of that kind. There is a classical character in the
literature, who did not believe in electricity. That woman was not confident
about electricity, in other words. What can you do about it? "Not to be
confident" is a mental category, not a scientific one. Prove that my
approach is incorrect. Challenge me not with this "hard to be confident"
stuff, those are empty words.

10. >"How do you see two clades within one dataset?"

I compose a haplotype tree. It shows separate, distinct branches, if they do
exist. Sometimes you can see them directly, by eye, but it takes an
experience to notice them. I would not recommend it.

11. Shallow branches are separate branches. They come from a recent common
ancestor. They should be tested separately.

12. >These leaves a lot of gaps".

Oh, c'mon. He, who wants, looks for possibilities; he, do does not want,
looks for excuses.

Please do not take it personally.


Anatole Klyosov

Dear Anatole,

Going through your last post which was now some time back...

>I will respond to the essence of your message later. I have not read your
message yet, because I bumped into the first paragraph (see above) and it
made me slightly astonished. ... I have noticed that every joke I made you
consider as if I speak before a High Court.

Please do not be distracted by my citation of your words. I do not doubt
that they were meant light heartedly. I cite them just because they are the
starting point for a reply.

>Second, please do not demand that I must (sic!) answer something when there
was no any real question.
>Responded to what?
>Can you specify your question?

I do disagree that I never asked any clear question. I think my question is
consistent, and is indeed based on trying to understand the conversation
between you and Ken which got stuck. (And I think you felt that others were
judging you too quickly without asking the questions behind their doubts?)
The most specific version at
02 was:-
>In my example above, which mutation happened first and is more ancestral,
on the following "slow moving markers":-
>*The 11 or 12 on the second marker?
>*The 16 or 17 on the third marker?
>*The 18 or 19 on the fourth marker?

> I have opened the link and immediately closed it. Did you really want me
to study those well over hundred of extended haplotypes

No, I suggested originally that it would not be necessary for my question.
Perhaps you didn't realize why. I do understand that the type of question
you get asked the most, and which you explain the most, is to look at a
bunch of haplotypes and estimate an age. We have also had such discussions.
However in this case I want to isolate the step which is prior to this part.

I asked only how you make a "haplotype tree" (your term), which as you said,
and maybe this is the point I need to understand better, is an essential
first step which many people get wrong. For example to John's example you
start by first saying the group had two clades, and in your articles you
also show trees. I would like to isolate this step and understand if you do
it in a special way. I think other people are also interested.

Concerning this first step, if you do it like the rest of us, in my opinion
it will very often just be a shot in the dark, whether you do it by eye or
by computer programme. If you do not, then I really want to learn the
method. I would be happy.

Above, I re-stated the question you thought I had not asked. Here was your
response first time round...
32 You wrote:- "This what can be shown on a haplotype tree. You cannot say
what is "more ancestral" since there are several ancestors there. THE
ancient "the most recent" common ancestor might not be even seen clearly
from the listed haplotypes. It can be deduced, though, by comparing base
haplotypes of several subsets, which can be identified on the haplotype
tree. Some of them can be poorly identifiable due to insufficient

So, going through the paragraph...

1. The haplotype tree will be shown in the haplotype tree.

2. If the statistics are insufficient, then identification can be

That's where we got to so far.

Best Regards


To contact the GENEALOGY-DNA list administrator, send an email to

To post a message to the GENEALOGY-DNA mailing list, send an email to

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
email with no additional text.

End of GENEALOGY-DNA Digest, Vol 5, Issue 114

This thread: