GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2006-08 > 1154467069


From: "William Hurst" <>
Subject: Re: [DNA] Sorenson mtDNA search engine bug
Date: Tue, 01 Aug 2006 17:17:49 -0400
In-Reply-To: <430.41a62f2.320107bc@aol.com>


Hi,

A couple of additions to Ann's response. As she said, it may be possible to
have an unpaired 524 insertion, but the current SMGF database has exactly
214 of each. The related 523,524 (or 522,522 in FTDNA-speak) are also
usually paired, but there is one more of the latter in the current database.
In my recent Genetic Distance reports, I count each pair as only one
mutation.

Bill Hurst

Ann wrote:

>
>In a message dated 8/1/2006 10:55:35 AM Pacific Daylight Time,
> writes:
>
> > There's a minor bug in the Sorenson search engine (it has a workaround).
> > It seems that when the search engine encounters a site with multiple
> > mutations, it treats them as all being part of the required search.
> > IOW when we search for 524.1A, you can't just search for no
> > variants--you have to have at least one difference to allow for 524.2C.
> > So to be specific when I searched on 499A and 524.1A, if I made the
> > allowed number of differences be zero, I get a sorry message. IfOTOH
> > one allows for one or 2 differences, one gets one difference hits on
> > BOTH 499A,524.1A and 499A,524.2C even though one didn't set 524.2C as a
> > search parameter.
> > JAL
> >
>
>In this particular case, it's best to enter both 524.1A and 524.2C as part
>of
>your search strategy. That region has several sets of "AC" in a row, and
>mutations (insertions and deletions of the "AC" motif) in that region
>almost
>always come in pairs. In fact, I can't recall off-hand seeing one without
>the
>other, although it's theoretically possible, like partial repeats in Y STR
>markers.
>
>Ann Turner
>
>
>==============================
>Search the US Census Collection. Over 140 million records added in the
>last 12 months. Largest online collection in the world. Learn more:
>http://www.ancestry.com/s13965/rd.ashx
>


This thread: