GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-07 > 1215471979


From: "Ken Nordtvedt" <>
Subject: Re: [DNA] J1 clusters
Date: Mon, 7 Jul 2008 17:06:19 -0600
References: <070720082235.14311.48729A4C00033BFA000037E7220074567290019C019704050E@comcast.net>


We're not talking 50 percent confidence intervals on your assumed average
mutation rates. We are talking the confidence intervals due solely to the
probabilistic nature of the mutation process itself which will be there if
we know the mutation rates precisely. There are not the Avogadro's number
order of magnitude of mutational events happening here in the typical tree
from founder to present sample population; so the n plus or minus square
root of n is significantly in play.

And you must explain your "calibration process". We can not go with
calibration by anecdotal example. If you are not "calibrating" the various
sum of marker mutation rates for your different sets of markers, what are
you calibrating?


----- Original Message -----
From: <>
To: <>
Cc: <>
Sent: Monday, July 07, 2008 4:35 PM
Subject: Re: [DNA] J1 clusters


> Ken Nordtvedt wrote:
>
>>I take it you mean k = total (sum of) mutation rates of haplotype markers?
>
> My response:
>
> Not exactly. It would be "too simple" and, unfortunately, not exactly
> correct due to still shaky figures for some loci/markers (see below).
>
> Basically, you are right. In kinetics, the rate of parallel reactions is
> equal to a sum of the "individual" reactions. If these reaction rates are
> first-order reaction rates, the "overall" reaction rate constant is equal
> to a sum of individual reaction rate constants.
>
> However, in the case of mutation rates in, say, 25-, 37- and 67-marker
> haplotypes one has to be damn sure that he/she adds correct mutation
> rates. I do not think we are there yet.
>
> That is why I CALIBRATEDd my mutations rates, empirically. For 6-marker
> haplotypes my figure is indeed very close to the sum of those six mutation
> rates. For 12-marker haplotypes it is pretty close, with an exception of
> one particular marker that we have discussed earlier. For 25-marker
> haplotypes there is about a 50% deviation between my "proven" calibrated
> average mutation rate, on the one hand, and a sum of 25 individual
> mutation rates (actually, 13 of them, since the first 12 work fine), on
> the other.
>
> 50% deviation would be strangely much for my calibration. I would suspect
> a few wrong figures for (mainly fast) individual markers. I do not insist
> that my mutation rates are ideal, but not a 50% deviation.
>
> Anatole
>
> -------------------------------
> 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: