GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-01 > 1105679152


From: Charles <>
Subject: Re: [DNA] Male Line Specific Y-STR Average Mutation Rates
Date: Fri, 14 Jan 2005 00:06:02 -0500
References: <1d7.3442b87e.2f18a27e@aol.com>
In-Reply-To: <1d7.3442b87e.2f18a27e@aol.com>


Ann,

OK let's play some show me.

You're playing your game with the same underlying mechanism...one
standardized deck of cards. The assumes the basic mutation rate for all
males is basically the same. The point of my theory is assuming
something different is going on in each male line's Y chromosome. Some
mechanism we don't understand.

Let's say that counting the hearts as mutations in your analogy. You got
a modal value of 3 with your experiment, or let's say that equates to an
average line mutation rate of .003.

What if every male line has a different deck of cards. Each deck would
have the same total number of cards as in a standard deck. But some
decks would have four more hearts and less other suits and some with six
more hearts in it and less other suits. And another male line has a deck
with four less hearts in it, etc. and another male line has a deck with
six less hearts in it, etc. Same number of total cards but a different
amount of hearts (potential mutations that can occur for the simulation
game you suggested) at random using your game. It is still a random
process within the specific male line, but since each male line has a
different deck of cards, one male with a deck with more hearts
(potential mutations in his games), and another with less hears
(potential mutations in his games), the number of hearts obtained
typically would not be the same for each game comparing a game played
with one standardized deck and other games played with different heart
amount containing decks.

Does that help you see what I am trying to say. Not all male lines are
playing from the same deck. Humor intended. :-)

I think that the Y chromosome has some unknown to us mechanism such that
results in a net average mutation rate that is different from male line
to male line. It will still be random within the male line but the
average mutation rate will be different from male line to male line. In
the hearts in the deck analogy, each male line is using a different deck
of cards, some with more hearts and another line has a deck with less
hearts.

Do this help you see my point? We need to think outside the box of
conventional wisdom on this subject. I don't think one rate fits all. I
think something is going on with each male line Y chromosome's average
mutation rate which makes the average rate vary from one male line
(surname project) to the other. And of course that underlying mechanism
is being passed down over the generations from father to son.

We need more data in an organized central location. You mention all the
random anecdotal comments on the List. Well I'm going to try and
consolidate the data so the power of synergy and numbers makes or breaks
my point. Same thing I did with all the random comments about the
DNAPrint test and East Asian vs. Native American. Let's pool all the
anecdotal evidence and see what falls out of the data when looked at as
a whole. My Log will collect the data. Whit and I and others will
analyze it. And if interesting things show up, maybe some population
geneticist and scientists will dig deeper into this hypothesis of mine.

I invite all surname project administrators who have a known common male
ancestor for their project participants and know or can triangulate the
ancestral haplotype, to join my new Log project.

Synergy at Work!

Charles Kerchner
http://www.kerchner.com/cgi-kerchner/ystrmutationrate.cgi
http://www.kerchner.com/dnamutationrates.htm

wrote:
> In a message dated 01/13/05 1:34:07 PM Pacific Standard Time,
> writes:
>
>
>>John,
>>
>>You keep focusing on the average and tend to always minimize the
>>differences observed between one male line surname project (y chromosome
>>observed average mutation rate) compared to another. But clearly some
>>surname projects are observing average mutation rates 3 or almost 4
>>times more than others.
>
>
> If a process is random, we SHOULD see some results which are higher than the
> "expected" value and some that are lower than the "expected" value, purely by
> chance. If we didn't see such a range, we'd be looking for a more mechanistic
> way of explaining how mutations happen. Perhaps the aspect that is bugging you
> is that the range seems so big.
>
> Charles, you strike me as a "show-me" kind of guy, so how about conducting a
> statistical experiment. Sit yourself down with a well-shuffled deck of cards
> and deal out a hand of twelve, then count the number of hearts. We would
> "expect" to see three, right? Put the cards back in the deck, reshuffle and repeat
> the process for 10 times. Then do another set of 10. Here are my results
>
> Trial Set 1 Set 2
> 1 2 3
> 2 3 2
> 3 3 1
> 4 4 4
> 5 3 4
> 6 6 2
> 7 3 3
> 8 3 3
> 9 5 3
> 10 4 3
> Total 36 28
> Average 3.6 2.8
>
> In Set 1, the results varied from 2 to 6 (a three-fold range) and in Set 2
> from 1 to 4 (a four-fold range). In the two sets combined, the range is 1 to 6,
> and if we kept on dealing out hands, eventually we'd find some more extreme
> "outliers," hands with 0 or 7 or even more. These are the kinds of cases we hear
> about on the mailing list, because they seem so odd, and they're the ones
> that stick in our memory. We don't hear so much about or recall the cases where
> the results are closer to "expected" values.
>
> If you combine the sets and construct a histogram of those results, you get
> something approaching the normal curve.
>
> Number of 1's: 1
> Number of 2's: 3
> Number of 3's: 10
> Number of 4's: 4
> Number of 5's: 1
> Number of 6's: 1
>
> We know all the hands were dealt from the same deck of cards, so we don't
> have to speculate about a "cause" for hands which had more or fewer cards than
> expected.
>
> Now if you came back and told me you got exactly three hearts on every hand,
> I'd accuse you of fudging the data or rigging the experiment. In fact, if you
> came back and told me the average for every set of 10 was exactly 3.0, I'd
> still suspect you of selectively reporting your data. Random results are more
> irregular than that.
>
> It's still worthwhile to flag cases that differ from the "expected" values
> and to keep an eye on them -- that's how discoveries occur. In fact, I want to
> think a little bit more about your case study. You're using an average
> yourself, you know, when you talk about the descendants of Adam Kerchner. Most of the
> mutations occur in one branch, the descendants of 1118.
>
> Ann Turner - GENEALOGY-DNA List Administrator
> Search or Browse the archives, Subscribe or Unsubscribe at
> http://lists.rootsweb.com/index/other/DNA/GENEALOGY-DNA.html


This thread: