Archiver > GENEALOGY-DNA > 2005-05 > 1116164772

From: Thomas Krahn <>
Subject: Re: [DNA] DYS 464, 459, and CDY Correlations and Deletions
Date: Sun, 15 May 2005 15:50:55 +0200
References: <00e301c55906$50b1c2b0$c6559045@Ken1>
In-Reply-To: <00e301c55906$50b1c2b0$c6559045@Ken1>


You can't blame the testing companies for giving "wrong" results. They
do actually the best they can. Multicopy markers are a very complex and
undiscovered issue. You cannot conclude from a single peak that people
have a deletion in a multicopy marker. From the experience of other
people's alleles it is still more likely that they have 4 copies at
DYS464 and 2 copies at the duplicated markers. But in deed they could
also have 8 or 16 copies. Biology is sometimes more complex than we think.

To be conservative, all testing companies should stop reporting
multicopy markers at all, because they could possibly give wrong
results. There are enough other single copy markers, that can be used.
We don't depend on them.

But I don't think that this is the right solution, because there's much
more behind multicopy markers than only a fascinating curiosity of
nature: The most interesting side effect from duplicated DNA regions is

Recombination effects are still poorly understood, but some known
parameters show us, that recombinational changes in DNA sequences may be
a great opportunity to learn more about our ancestry. The frequency of
recombination is slower than STR "slippage" mutations, but quicker than
SNP mutations. That means, that if we learn how recombinational DNA
changes work, we could find out more about the "time-gap" between STRs
and SNPs.

I recently posted some interesting facts about DYS464 patterns with
separated allele calling. Meanwhile I found a (very) few exeptions, that
show that the 3 x "C" & 1 x "G" rule is not 100% strict. This shows us,
that recombination effects really happen.

Just to give an imaginary scenario how a recombination change _could_
Let's say we have 4 differrent DYS464 alleles: 13,14,15,16. If we have
some strange mutation in a coding region that is associated with the
"14" allele, the man would be infertile. But by chance one stem cell in
the germline has been able to to "repair" the defect by simply copying
one of the "backup" copies, let's say the region with the allele 15. Now
the man is fertile again and can be the founder of a whole population.
His DYS464 fingerprint is now reduced by one variety (13,15,15,16), but
no STR slippage mutation has happened. A similar reduction of variant
DYS464 alleles could have happened in haplogroup N, but a bigger region,
covering 2 or 3 DYS464 alleles has been involved. Then there would be
still 4 alleles in DYS464.

I hope recombination will soon be better understood. This gives us a new
tool to discover human migrations and more...

Thomas Krahn

Ken Nordtvedt wrote:
> There are some very interesting correlations between duplicates of reported repeat lengths at DYS464a,b,c,d with DYS459a,b and CDYa,b throughout all Yhaplogroups.
> Whenever DYS464a,b,c,d is reported to be of the form X,X,Y,Y then there is a higher than normal chance for DYS459a,b to be reported as Z,Z and also CDYa,b to be reported as W,W.
> If these three markers had their multiple alleles physically located close to each other, then deletions or suppression mechanisms which eliminated an allele for one of these markers would tend to eliminate it for the others, thus explaining the contagion of the duplicate reported lengths. I have checked databases of I1a, I1b, I1c, and N haplotypes, and this is happening in all of them.
> What bothers me is why these markers are being reported by the testing companies with their full compliment of allele lengths (4 for 464, 2 each for 459 and CDY) if deletions or blockages of some kind have occured which eliminated alleles from these markers' sets?
> With 4 copies of one of these markers, and 2 each of the other two, there are all kinds of ways one could locate these things close to each other with different resulting patterns of correlated losses of alleles for the 3 markers, depending on the location of the deletion(s) or other suppression methods. Since elimination of two of DYS464's copies occur now and then, and that is apparently the modal situation for haplogroup N, would not it be possible to sometimes have both copies of DYS459a,b and/or CDYa,b lost and therefore entirely missing in some haplotypes? Are these markers missing more often than other markers?
> I have a repeat of 14 among my 4 DYS464 repeats, and both FTDNA and DNA Heritage have reported this same result. A statement from each as to how they conclude there are two repeats of length 14 in my ydna would be of interest.
> Although there are so many interesting things to speculate about at this point, the main purpose of this note was to mention the interesting correlations between the three special markers indicated above. The canon has been that markers are independent of each other.
> Ken
> ==============================
> Search the US Census Collection. Over 140 million records added in the
> last 12 months. Largest online collection in the world. Learn more:

This thread: