GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2002-08 > 1028562579


From: "James Reynolds Hull" <>
Subject: Re: [DNA] thanks
Date: Mon, 5 Aug 2002 09:00:26 -0700
References: <5.1.0.14.0.20020805062348.045078b0@mail.verizon.net>


John,

I agree with your rational for markers. One slight addition I would add for
12/10: Wait until a few more have tested. I had a similar case, and by
waiting for additional testing results, solved the problem and developed an
independent haplotype.

Jim Hull
----- Original Message -----
From: "John S Walden" <>
To: <>
Sent: Monday, August 05, 2002 4:34 AM
Subject: Re: [DNA] thanks


> All
> Reviewing the document at
>
http://www.journals.uchicago.edu/AJHG/journal/issues/v66n5/991495/991495.htm
l
> For us I believe one of the keys included is this quote
> "Mutation-rate estimates are necessary in such paternity testing, to
> assess the possibility of a potential false exclusion."
>
> What I think this says is that in the absence of a paper record and you
> wish to be really confident you make the right
> decision or excluding/including when the tests are close then the 25
marker
> test should be used.
> Here is an example of what I think they are saying to us.
> If we do a 12 marker test and we have a 5 generation time frame (4 births
down
> each side of the tree) then:
> Markers
> Tested/Match What to do
> 12/12 Probable match with out paper trail you must expand to
> 24/25 marker test
> With a paper trail accept the match and test more
> people.
> 12/11 Same
> 12/10 Likely a mismatch upgrade to avoid the 1 out of 20 case
> of a false exclusion.
> 12/9 Very likely a mismatch spending the money will
confirm it.
> This is a 100 to one shot so only spend the money
> if you
> want to go from 99% to 99.999% certain.
> 12/8 (and less) Forget it there is NO match here.
>
> So in the last two cases one can save the cost of testing to 24/25
markers.
>
> If the number of generations involved goes up to 10 (9 births down each
> side of the tree)
> then the 100 to one breakpoint is between 4 and 5 mismatches
> That is at 3, or less, test more; at 4 or 5 think about it; and at 6
> differences there is no family match.
> [If the people we are testing are born about 1950 and at 25
> years/generation takes us
> back to the early 1700s so this looks to me to be our more usual case.]
>
> That is the way I calculate the numbers.
> For Surname project managers would that make sense from what you have
seen?
>
> Thinking further on this 6 (more) out of 12 differences sounds like a lot
to me
> I think I will look through the Surname studies and see how often this
happens.
> Maybe is it so rare we should just forget about it an figure we should
> spend the money and
> be sure right from the start.
>
> John W
>
>
>
> At 11:58 PM 8/4/2002 -0400, you wrote:
> >Thanks to Ann and Barbara.
> >
> >I've located
> >
>
>http://www.journals.uchicago.edu/AJHG/journal/issues/v66n5/991495/991495.ht
ml
> >
> >and I think it is saying, "Upgrade to a 25 marker test!"
> >
> >Grant
>
>
> ==============================
> To join Ancestry.com and access our 1.2 billion online genealogy records,
go to:
> http://www.ancestry.com/rd/redir.asp?targetid=571&sourceid=1237
>
>

This thread: