GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2009-12 > 1260800468


From: Al Aburto <>
Subject: Re: [DNA] R-U152 and R-L21 on the European Continent
Date: Mon, 14 Dec 2009 06:21:08 -0800
References: <200912131920.nBDJK3Bq025158@mail.rootsweb.com>
In-Reply-To: <200912131920.nBDJK3Bq025158@mail.rootsweb.com>


Tim,
One thing is that DYS388 is not a slow mutator in Y-DNA J.

Another is what do we do with TMRCA between two haplotypes where
generally (FTDNA & others) use much faster rates? A one marker mismatch,
at say DYS437, (out of 24 slow markers) indicates a TMRCA of ~12000
years ago or less with 95% probability using a mutation rate of 0.0004
mutations per generation and 30 years per generation. What would Dr.
Walsh and FTDNA say about that?

But you say it is OK to use the fast mutators for the younger clades :-(

I'm not sure what to think about all this (I sort of gave it all up
quite a while ago) :-(
Al

> Tim Janzen wrote:
> Dear Al,
> Yes, for the "24 Slow" option, I throw away the "43 non-slowest"
> mutating markers. Bear in mind that 17 of these markers are discarded at
> the onset because they are multi-copy markers. We discussed the rationale
> for this on the list 1½ years ago when Ken and I were first developing
> intraclade and interclade TMRCA estimator programs in Excel. The other 26
> markers are thrown away simply because their variance is saturated (or at
> least partially saturated) when you are looking at TMRCA estimates for old
> haplogroups and subclades. You can retain all 67 markers in intraclade and
> interclade TMRCA calculations and this is doesn’t create significant
> problems when calculating TMRCA estimates for young subclades. However,
> retaining all 67 markers for interclade TMRCA estimates for old haplogroups
> and subclades creates the same problem that including all of the 50
> single-copy markers does in that the interclade TMRCA estimates are
> consistently too low for old haplogroups and subclades. I haven’t been
> reporting to the list any interclade TMRCA estimates using 67 markers for
> about 1½ years because of this problem. For comparison purposes you might
> be interested in knowing that the interclade TMRCA estimate for haplogroups
> A and B is 46,848 years when using all 67 markers. This compares to 35,245
> years I got for 50 markers. In any case, we all know that 46,848 years is
> also lower than the true TMRCA for haplogroups A and B.
> The average mutation rate for the remaining 24 markers is .00040.
> Zhivotovsky et al (2004) use .00069. The 24 markers in my "24 Slow" option
> are listed below.
> Sincerely,
> Tim
>
> 472
> 578
> 426
> 455
> 454
> 641
> 436
> 490
> 450
> 388
> 594
> 640
> 531
> 617
> 492
> 392
> 568
> 590
> 438
> 537
> 393
> 565
> 487
> 437
>
> -----Original Message-----
> From:
> [mailto:] On Behalf Of Al Aburto
> Sent: Sunday, December 13, 2009 9:45 AM
> To:
> Subject: Re: [DNA] R-U152 and R-L21 on the European Continent
>
> Hi Tim,
> So for the "right mix" ("sweet spot") we throw away the "43 non-slowest"
> mutating markers, giving more sensible TMRCA results. This means the
> Y-chromosome, as a whole, having some sort of average mutation rate, is
> better represented by these 24 slowest mutating markers. I think I
> asked this before, but is your average mutation rate for these 24
> slowest markers similar to what Zhivotovsky et al (2004) use?
> Al
>


This thread: