Archiver > GENEALOGY-DNA > 2003-11 > 1069804430

From: Charles <>
Subject: Re: [DNA] Measuring Genetic Distances -- marker [DYS464] is a specialcase and controversial marker
Date: Tue, 25 Nov 2003 18:53:52 -0500
References: <>
In-Reply-To: <>


Respectfully, as you know, DYS 464 and also DYS 389i and DYS 389ii are
two special case markers which need special attention to determine the
actual difference for those respective markers, before one even attempts
to compute the overall Genetic Distance for one's haplotype compared to
another's haplotype in a surname project, whether one uses the overall
Genetic Distance prediction method using the simple sum of the
differences or the prediction method of the sum of the squares of the
differences at each marker advocated by some in this forum. DYS 464 is
an exceptionally controversial one. Because of DYS 464's confusing
nature some companies have elected not to even use DYS 464 because of
the potential confusion it can introduce. I contacted and asked Dr.
Walsh to join the List in order to debate and discuss the math and
statistics (for each advocated method of summing up the differences for
comparing two haplotypes) with those advocating the "sum of the squares
method" once the differences for each marker are determined. I wonder if
posing that question about DYS 464 at this point in this discussion may
confuse and or lose the original math question thread, for all the
newbies reading and not so newbies too, and take us into another
confusing issue, i.e., the unique issues of difference counting
regarding marker DYS 464. Whatever the method used for determining the
difference at the special case markers DYS 464 a b c and d, which may be
more in the realm of Dr. Hammer's expertise in FamilyTreeDNA's advisory
staff than Dr. Walsh's, does one then sum them up or square them first
and then sum them up. I believe Dr. Walsh's method is to use the simple
sum them up method as being most predictive of the actual genetic
difference. I was hoping to hear more from those who disagree with that.

Can those in this forum who are mathematicians and statisticians please
comment on Dr. Walsh's email and website in regards to using the simple
"sum of the differences" prediction method for the rest of the haplotype
markers when comparing two haplotypes and leave the special case markers
out of it for the time being.



> In a message dated 11/25/03 1:45:52 PM Pacific Standard Time,
> writes:
>>The IAM model scores
>>all three of these as 23 matches and 2 mismatches, while the stepwise model
>>weights the actaul number of steps off when considering the time to MRCA.
> Sorry, I just sent out an empty message in the course of searching for an
> example. I more-or-less understand the general principle, but I'm not sure it's
> being handled in the best way for DYS464a-d. The fact that we don't know the
> actual value for each physical marker confounds things. For example, take these
> values:
> 15 16 18 18
> 13 15 16 16
> If you compare a to a, b to b, c to c, and d to d, the Genetic Distance would
> be 2 + 1 + 2 + 2 = 7. The FTDNA "Genetic Distance" report for the DYS464a-d
> portion amounted to just 2, because both results have a 15 and a 16 in common,
> but no matches on the remaining two markers. I think that must be a gross
> underestimate of the "real" Genetic Distance.
> Ann Turner - GENEALOGY-DNA List Administrator
> Search or Browse the archives, Subscribe or Unsubscribe at
> ==============================
> To join and access our 1.2 billion online genealogy records, go to:

This thread: