Archiver > GENEALOGY-DNA > 2003-11 > 1068521518

Subject: [DNA] SNP test for R1a/R1b in view of Kittler paper
Date: Mon, 10 Nov 2003 22:32:02 EST

We recently discussed a paper by Kittler, which showed
that DYS385a/b can give misleading results because we
can't tell which marker is which. You can search the archives
for more discussion. This URL ends in html:

The following message was posted on the FTDNA message
board. I'm quoting it with Don Potter's permission. Don also
sent me his actual values, which I'm putting in a table below
his message.

------- begin message from Don

My results suggested that I was haplogroup R1a and I paid
the extra $100 for the SNP test to find out I was actually
R1b. With this possibility I would think it best to spend
the extra money for the test to be certain.

My test showed no exact matches, four one step mutations
(all R1a), twelve two step mutations (also all R1a), thirty
six three step mutations (31 R1a and five R1b), and a bunch
at the four step mutation point with over half of them
being R1a which I am not.

What accounts for this? Well, I called FTDNA and spoke with
both Max and then Bennett who explained this all to me.
Basically they have what he called a haplogroup calculator
that allows them to take a "best guess" (which is usually
right) at which haplogroup you belong to. However, they or
you for that matter can not be sure as to the correct
haplogroup you belong to without taking the SNP test for
the suggested haplogroup.

As to why I had so many matches at R1a and I'm R1b was also
explained to me as what is referred to as convergence. I am
definitely R1b as the SNP test showed me positive for P25
which is the identifying marker for R1b but my changes down
through time have shifted to look more like R1a than what
one would expect of an R1b person. So as you can see the
SNP test is the only way to be certain. Hope this helps.

Don Potter

--------- end message from Don

When we look at the modal haplotype values on Dennis
Garvey's site, we see that Don matches 9 out of 12 for R1a
and 8 out of 12 for R1b. So far, he seems to be a better
match for R1a.

But in the Kittler paper, where they could detect which value
was which for DYS385a/b, the modal value for R1a was 11-14
and for R1b it was 14-11. If we could tell which was which,
Don's matches for R1a would probably drop to 7 out of 9,
and R1b would be the "winner." The ancestral haplotypes for
R1a and R1b in Kittler's paper are the last two columns in this
table. Note that they are pretty good matches for the modal values
on Garvey's site.

R1a R1b Don P R1a-K R1b-K
DYS393 13 13 13 13 13
DYS390 25 24 25 25 24(23)
DYS19 15 14 14 ? 14
DYS391 10 11 10 10 11(10)
DYS385a 11 11 11 11 14 <----
DYS385b 14 14 14 14 11 <----
DYS426 12 12 12
DYS388 12 12 12
DYS439 10 12 11
DYS389i 13 13 13 10* 10*
DYS392 11 13 11 11 13
DYS389ii 30 29 29 17* 16*

* The paper used Oxford Ancestor's nomenclature
for DYS389. 10-16 is the same as 13-29, and 10-17
is the same as 13-30.

As far as I can tell, FTDNA bases its "best guess" on
just the first 12 markers (still more than the Kittler paper
or any other database I know of).

If we continue comparing Don's values with Dennis'
table for the remaining markers, he has many more
matches with R1b.

R1a R1b Don P
DYS458 15 17 18
DYS459a 09 09 9
DYS459b 10 10 10
DYS455 11 11 11
DYS454 11 11 11
DYS447 23 25 25
DYS437 14 15 15
DYS448 20 19 19
DYS449 32 29 31
DYS464a 12 15 15
DYS464b 15 15 15
DYS464c 15 17 16
DYS464d 16 17 17

I hope I've transcribed everything correctly: I was switching rows and
columns, and changing Kittler's Hg1 to R1b, etc., so let me know if you spot any

So what's the bottom line?

1) You can't always determine your haplogroup from your haplotype.

2) Don's results are a few steps away from the modal haplotype, so his case
may be more difficult than usual.

3) The DYS385a/b marker may confound the interpretation (my assumption, based
on Don's haplogroup test).

4) It may be useful to compare your values on the full 25-marker test with
the modal values on Dennis' web page.

5) Whether Don is R1a or R1b probably makes no difference for his research
into a genealogical time frame, where he can use all 25 markers to compare his
results with other Potters.

Ann Turner - GENEALOGY-DNA List Administrator
Search or Browse the archives, Subscribe or Unsubscribe at

This thread: