GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2001-11 > 1005036277


From: Patrick Guinness <>
Subject: [DNA] Loci used in Guinness study
Date: Tue, 6 Nov 2001 08:44:37 +0000


We were going to aim for 10-12 loci, and then the academics said they
would like to try some new ones. So 17 is an experiment but it will
still show similarity and distance. Looking at 2,000 Irish donors is
a slow process.

Some say that DYS434, 435 and 436 don't change much.

The proposed results database is a great idea, but must allow
latitude for studies which have more or less loci. For example, in
the Jefferson-Hemings study (let's ignore the politics and whether he
was the father, or not, for a minute), the usual DYS- 19, 388, 390,
391, 392 & 393 are there, with others.

In case anyone would like to know if Jefferson was their ancestor as
well, his cousins' microsatellite results were reported like this -

Locus DYS
19___388___ 389a___389b___389c___389d ___390___391__392__393__DXYS156y

repeats
15___12_____4______11_____3______9_______11___10___15___13___7


The 389 is split into four sections, a b c & d, where DYS389D = DYS389i
and 389ii = A+B+D. The C is short and mutates little, so is ignored
in the i/ii system.

so the four DYS389 elements above should read, in the usual parlance,
389i=9 repeats
389ii=24.
As we're then double-counting the D, Oxford Ancestors reports
'389ii-i' which would show 15.

Why was the Jefferson DYS389 split into 4 in the first place? Because
the report was done in 1998 which was almost back in the Middle Ages
in this field.

DXYS156y has 11 repeats.

Also MSY-1 was included which requires at least 6 digits to record.
It's the one that mutates quickest, and I would recommend it for
family studies where most of the testees have a known tree. We're not
using it because we don't need to.


So allow for maybe 50 loci, and part-matches. In a few years we will
have someone with 26 loci and another with 36, who overlap on 25 in
common, etc. If 50 loci = one likely change in 10 generations, that's
pretty close to a realistic limit back to the end of most 'paper
trails.'

Definitely leave a space for an email address / web site and let them
get on with it.
Have an on-screen entry form so you don't have to do any transcribing.

Patrick Guinness

This thread: