Archiver > GENEALOGY-DNA > 2009-07 > 1247510708

From: "Ken Nordtvedt" <>
Subject: [DNA] Mutations
Date: Mon, 13 Jul 2009 12:45:08 -0600
References: <><><><><000801ca03c7$90c8bb90$6400a8c0@Ken1><>

----- Original Message -----
From: "Sasson Margaliot" <>

I have changed the title. I get fed up with topics of general concern being
discussed under R1b-titled threads. Down with imperialism.

> Ken
> If. for example, the mutation rates at some period in the past were
> faster,
> it would explain
> why the mutation rates of slow markers are more significantly
> under-measured. [[[ Please explain! The variances we get for all STRS,
> using the databases, cover identical time depths. Only if the fractional
> rate of change of mutation rates of STRs as we move into the past were
> different would the slow markers' versus fast markers' rates change
> relatively to each other. Take 393 rate divided by 462 rate as example.
> Let's just say the ratio today is about 15. If the ratio of these rates
> was 10 2000 years ago, that would screw things up and require refined
> model. But if that ratio stayed 15 as you moved into the deep past, but
> the two rates sped up, then you would not see estimated ages depend on
> which STRs you used. ]]]

[[[ Also, in the Chandler/SMGF method, where they are in effect counting
mutations which happened in the past, all STRS face the same profile of time
depth in their counts. So relative mutation rate determinations can divide
out that unknown time depth profile. ]]]
> It may also be the case that it is simply more difficult to measure the
> rate
> of change for super-slow markers. [[[ Since many of the slow markers are
> on the 38-67 panel, which has fewer haplotypes in the database, we have
> poorer measurement of their rates. But that by itself does not explain
> bias in their rates --- just uncertainty of what each rate is; some would
> be higher and some lower than measured. ]]]
> Whatever the explanation, maybe we we should look at the collected
> STR-data
> as the measurement of relative mutation rates - rather than of ages of
> various branches. [[[ ???? Don't understand. ]]]
> Will the calibrated rates for medium and slow markers will produce
> reasonably uniform age estimates? [[[ Again, not sure what you mean? ]]]
> Sasson

This thread: