GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-04 > 1270715717


From: "Sandy Paterson" <>
Subject: Re: [DNA] FW: Questions about alternate means of searchingDNA Genealogy
Date: Thu, 8 Apr 2010 09:47:09 +0100
References: <000001cad64e$0eb3ad90$2c1b08b0$@com><REME20100408001410@alum.mit.edu>
In-Reply-To: <REME20100408001410@alum.mit.edu>


You wrote

>I have to wonder, though, how many other candidate cousins have been
>missed by failing to do the Right Thing, which was and still is: look
>for the closest matches first.

My response :

I'm afraid I can't answer that. I first became aware of Jonathan Clayborn's
quest to find his ancestors when he posted on 29 March. His posting didn't
include his full 25 marker haplotype, only the markers that he felt were
different to typical R1b.

You also wrote :

>Not likely. I conducted an experiment this evening with a sample of
>12,000 entries from Ysearch and 29,000 from DNA projects. Among the
>Ysearch entries, the median genetic distance to the closest neighbor
>among the 12,000, regardless of surname, was 1, and the mode was 0.
>However, many of them were not so fortunate. For example, about 300
>"lonesome souls" had nearest neighbors with a genetic distance of 6.
>Of those 300 not-really-neighbors, about 100 had surnames with
>projects included in the collection of 29,000 project members. Of the
>100 now-hopeful lonesome souls, nobody had a close match in the
>targeted project, except for seven who turned out to have the same
>surname as the project all along.

My response :

You sound a bit like Eugene Fama. He personally was unable to find a way of
beating the stock exchange, so he concluded that no one else could. And so
the efficient markets hypothesis was born.

I'll explain by using Stroup/Dougherty as an example. Stroup is kit number
24073 so he was tested some time ago. By that stage, there were no close
matching Dougherty haplotypes. Then along came Dougherty kit number 31706,
gd of 6 from Stroup (all of this is for 67 markers), 2 off-modal matches. A
clue, perhaps, but nothing to get excited about. Then followed, in kit
number order (and therefore presumably in date order) :

GDOff-mod matches

3818071
4280675
4564154
4681962

Eventually, though :

7531026

That's the principle, a fairly obvious one, I'd have thought.

Do you exclude DNA projects below a certain minimum size?


Sandy



-----Original Message-----
From:
[mailto:] On Behalf Of John Chandler
Sent: 08 April 2010 05:38
To:
Subject: Re: [DNA] FW: Questions about alternate means of searching DNA
Genealogy



This thread: