GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2011-10 > 1319068219


From: "Kenneth Nordtvedt" <>
Subject: Re: [DNA] [yDNAhgI] Fw: TMRCA Calculations
Date: Wed, 19 Oct 2011 17:50:19 -0600
References: <8D2C55FEDCB942CC9298C8C5657CD1C2@kenPC><CACPdyDm3M4ru3FDj4xMGjY=tSv5JtmZNLV1OArgyy7_6utMPjg@mail.gmail.com>
In-Reply-To: <CACPdyDm3M4ru3FDj4xMGjY=tSv5JtmZNLV1OArgyy7_6utMPjg@mail.gmail.com>


-----Original Message-----
From: Terry

Re-reading your comments Ken, you appear to make a fundamental distinction
between "coalescence age" and "TMRCA". How are we to interpret your earlier
postings?

The definition of TMRCA ("time-to-most-recent-common-ancestor"), applied to
any population, seems clear to me otherwise.

[[ Yes, there is a clear distinction. TMRCA is time back to the most recent
common ancestor, i.e. the founder. That's a unique event. Coalescence age
is sort of a group property of a tree --- average tmrca of the N(N-1)/2
pairs of haplotypes you can take from the set of N final haplotypes. They
are not only conceptually different, they are estimated by different types
of variance. KN ]]

a_____________e
b_____________|_______f_____
c_______g_____________|
d_______|

age back to f is the tmrca for haplotypes a,b,c,d

age back to e (a,b MRCA) plus age back to f (a,c MRCA) plus age back to f
(a,d MRCA) plus age back to f (b,c MRCA) plus age back to f (b,d MRCA) plus
age back go g (c,d MRCA) divided by 8 is my working definition of
coalescence age for a,b,c,d. Others might divide by 6 instead of 8.

This normalization issue could be discussed further if folks desire. It is
a technical issue; not an important conceptual one. If forced to by higher
powers, I could live with either normalization.]]

This thread: