Archiver > GENEALOGY-DNA > 2010-02 > 1266633229

From: "Anatole Klyosov" <>
Subject: Re: [DNA] TMRCA assessments
Date: Fri, 19 Feb 2010 21:33:56 -0500
References: <>

>From: Robert Stafford <>
>Did you calibrate the rate constants as rate per haplotype per year and
>divide by 25 to get the per-generation value or did you calculate the
>latter directly?

O.K. One more time.

Suppose, Tutankhamun had issue, and now you manage to identify 1820 of his
descendants. You got their 1820 of 25 marker haplotypes, and saw that eight
of them are identical to each other, while others are all different and
mutated compared to those eight.

Since you are smart and thoughtful, you are able to figure out that those
eight haplotypes are "base" haplotypes, that is they allegedly represent the
haplotype that Tut himself had. The ancestral haplotype, that is. However,
being smart and thoughtful, you are not sure that the haplotype is indeed
the ancestral one. You would like to verify it. Since you surely know that
King Tut died 3333 years ago, you decided to count it as 134 generations ago
with 25 years per generation (3350 years, rounded).

To make sure that those 8 identical haplotypes do indeed represent the base
haplotype, you employ the logarithmic method. You write: [in(1820/8)]/0.046
= 118 generations w/out a correction for back mutations, or (look up at my
table in JoGG) 134 generations with the correction.

Voila. You got exactly 134 generations, or 134x25 = 3350 years. It means,
that those 8 haplotypes are the base haplotypes. So far, so good.

You ask - why did I use 0.046? What is that? My answer - because it is the
average mutation rate constant for 25-marker haplotypes, as a number of
mutations per haplotype per generation of 25 years.

You just have performed the calibration. You saw that with the mutation rate
constant of 0.046 mutations per 25-marker haplotype per generation of 25
years you got the right timespan of 3350 years. In fact, Tut dies 3333 years
ago, but it is only 0.5% off. I hope you are in peace with such a precision.

However, you want to check calibration with mutations as well. Someone has
said here that nobody can count mutations, so you are in a kind of limbo.
O.K., let's take a look at those 1820 haplotypes. They all contained 9879
mutations, if to count them from that very base haplotype.

You have: 9879/1820/0.046 = 118 generations w/out correction for back
mutations, or (look up in the Table) 134 generations corrected for back
mutations, that is 134x25 = 3350 years to King Tut.

You just have performed the calibration again. Now, please remember that
figure: 0.046 mutations per 25-marker haplotype per generation of 25 years.
This is the average mutation rate constant for 25-marker haplotypes.

If you repeat the same procedure with 12-, 37-, 67- marker haplotypes, as
well as with ANY haplotype format, you get mutation rate constants for all
of them. If you pick other events for calibration, historical events or
genealogical ones, you will see that the mutation rate constants stay about
the same.


Anatole Klyosov

On Wed, Feb 17, 2010 at 5:56 AM, Anatole Klyosov

> More on the subject.
> A calibration requires two things: (1) knowing (or assuming) a time span
> to
> the event, being considered for the calibration, and (2) a total number of
> mutations that occurred in the haplotype dataset compared to the base
> haplotype. Oh, yes, and a proof that both linear and the logarithmic
> methods give the same value (for that you do not need a mutation rate
> constant),
> that is M/G = ln(G/n). Here M = a number of mutations in the dataset, G is
> a number of haplotypes in the dataset, and n is a number of base
> haplotypes
> in the dataset.

This thread: