GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2009-03 > 1238296568


From: (John Chandler)
Subject: Re: [DNA] Multi-step mutation rates.
Date: Sat, 28 Mar 2009 23:16:08 -0400
References: <000301c9afec$ee30fba0$ca92f2e0$@rr.com>
In-Reply-To: <000301c9afec$ee30fba0$ca92f2e0$@rr.com> (ewade@cfl.rr.com)


Eldon wrote:
> Their data set contained 118,866 parent-offspring transmissions from 59,433
> parent-offspring triplets. That study reported that the maximum-likelihood
> estimate of the exponential decay rate parameter {lambda} was 1.06 (Table
> 1). They said that suggests that 65% of mutations were of step size 1, 23%
> of size 2, and the remaining 12% of step size greater than 2.

That's not what has been seen in the YDNA studies of father-son pairs.
I don't have the numbers in front of me, but, as I recall, more than
90% of the Y mutations are 1-step.

> Is it possible to extrapolate the separate percentages for mutations of step
> size 3 and step size 4?

Yes, their model is just an exponential. They are predicting 8% to
be 3-step and 3% to be 4-step. But, again, that's from an autosomal
study, not a Y study.

John


This thread: