GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2002-07 > 1027049040


From: "John F. Chandler" <>
Subject: Re: [DNA] RFLP siblingship tests
Date: Thu, 18 Jul 2002 23:24 EDT
In-Reply-To: DNACousins@aol.com message <120.1365ffc7.2a68db75@aol.com> of Thu, 18 Jul 2002 21:03:13 -0600


Ann wrote:
> > Allele Values
> > --------------------------------------------------------------------
> > marker ME Sib1 Sib2 Sib3 Sib4 Sib5
> >
> > D2S44 3.11 3.11 1.58 3.11 1.75 1.58
> > 1.58 1.75 1.28 1.75 1.58 1.28
> >
> > D4S163 4.49 5.11 4.49 5.11 4.49 5.11
> > 3.20 4.49 1.68 4.49 1.68 4.49
> >
> > D6S132 2.73 2.73 3.21 2.73 2.73 3.21
> > 1.24 2.73 1.24 1.24 2.73
> >
> > D7S467 6.58 5.20 5.20 5.20 6.58 6.58
> > 4.44 3.17 3.17 3.17 5.20 5.20
> >
> > D10S28 3.52 4.13 3.39 4.13 4.13 4.13
> > 1.74 3.52 1.51 3.52 3.52 3.52

> I approached this more like a logic problem. It seems to me that
> markers D2S44, D4S163, D6132, and D7S467 are all fully consistent with all
> six of you being full siblings.

Actually, there's also a problem with D2S44. True, there are only 4
alleles represented in the 6 offspring, but they are distributed
awkwardly. Consider 3.11, which occurs in ME, Sib1 and Sib3. If
everybody is a full sibling, then the 3.11 in each case had to come
from the same parent (let's call him or her "Pat" and the other parent
"Frankie"). Since each child gets one allele from each parent, that
means ME got a 1.58 from Frankie, while Sib1 and Sib3 each got a 1.75
from Frankie. That means Frankie had both 1.75 and 1.58, and the
remaining allele 1.28 must have been Pat's. Sounds ok so far, but
there's a problem: Sib4 wound up with 1.75 *and* 1.58, which we
determined both came from Frankie. That's a contradiction, so the
original assumption must be wrong.

We run into a similar problem with D4S163. Every child there got a 4.49
(say from Pat). That means the other *three* alleles all had to come
from Frankie. Another contradiction.

D6S132 may be ok, but there's one number missing. If that number is
either 1.24 or 3.21, there are only three alleles for this one, and
we can make a consistent picture. Otherwise, it's the same problem
all over again.

D7S467 seems to be truly ok at least.

Still, there's no way to reconcile the results with the full-sibling
theory. If you leave out "ME", then the contradictions all vanish.
Dropping Sib2 would cure the five-allele problem on D10S28, but would
*not* help the others. Dropping Sib4 would cure the inconsistency on
D2S44 but not the others. The simplest explanation, then, is that
"ME" is the one who doesn't fit.

John Chandler

This thread: