GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2012-03 > 1332454101


From: Jim Bartlett <>
Subject: Re: [DNA] Phasing 4th Cousins
Date: Thu, 22 Mar 2012 17:08:21 -0500 (CDT)


On 03/22/12, John Chandler<> wrote:

Jim wrote:
> Wouldn't each of our ancestors have just two atDNA-segment quilts (one
> paternal, one maternal) to cover each of their 22 chromosomes -
> regardless of the number of spouses? Same for each child?
No, in general, each ancestor has an undivided half-interest in the
four quilts common to the ancestor and his/her spouse. Since you are
proposing to populate the quilts by segments found in matches with
other testees, and since, in general, the identified common ancestors
come in pairs, you can't separate the two.
John Chandler


John - I might not have worded my statement properly. When I spoke of "each
of our ancestors" I was thinking of each of them as a single individual.
Nothing to do with their spouse(s). Each one got their atDNA from their two
parents. When you look at the segments attributed to a particular
Chromosome, say Chr3, there are segments from one parent which would map
across Chr3 - including some large segments; and there would be segments
from the other parent which would map across Chr3 - including large
segments. The large segments from each parent could look the same, they
could significantly overlap - but the segments of each parent would map Chr3
differently. So when I look at a large (IBD) segment shared by me and my
atMatch, that IBD segment could have come from my father or mother. Say that
segment was 20cM on Chr3 from 30 to 50m [made up numbers]. I could have
another segment 15cM on Chr3 from 35 to 50m with a different atMatch. This
second segment could be from the same or a different parent. Both of these
segments could be from the same ancestor of one parent or the other. Or the
segments could be from different ancestors, in which case one of these
ancestors would be an ancestor of one parent, and the other ancestor would
have to be an ancestor of the other parent. This is because there are only
two quilts possible for Chr3 - one from the ancestors of one parent, and one
from the ancestors or the other parent.

Another way to look at it is by taking one site on a Chromosome. Say it is
AG. The A is from one parent, the G is from the other. The A would be part
of one segment - maybe large, maybe small; the G would be part of another
segment from the other parent, also of any size. As you move along the
Chromosome, site by site, you attribute one letter to one parent, and one to
the other (classic phasing, as I understand it). At the end of the
Chromosome you've attributed a letter to each of the parents - you've
completed the two quilts for that Chromosome. As we look at our atDNA
results we see segments on each Chromosome, that are shared with an atMatch.
The large segments are deemed to be IBD - they come in tact from a single
ancestor. Each of these large segments must be a part of one of the two
quilts for that Chromosome. And all the segments we see in our atDNA results
must fit on, or be a part of, one of the two quilts for each Chromosome.

As I think about your statement, I can see four quilts in our Common
Ancestors (two quilts for the husband and two quilts for the wife), and of
course we don't know which of these quilts provided the IBD segment that
came down to me and my atMatch. But I'm not trying to determine my
ancestor's quilts; I'm trying to determine my own! If I can determine an
IBD segment, shared by two other distant cousins, that goes back to a
distant ancestor, then I have identified that quilt piece on my own
Chromosome. That quilt piece is then my road sign back to that ancestor.
Eventually, I'll be able (in theory, and with enough atMatches) to determine
the boundaries of my Chromosome quilts, and attribute various (not
necessarily all) paternal ancestors to various segments on one quilt per
Chromosome, and ditto for Mom's quilts.

Jim

This thread: