GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-05 > 1211513075


From: Thomas Gull <>
Subject: Re: [DNA] S21/S28 Split+m223 stuff
Date: Thu, 22 May 2008 23:24:35 -0400
References: <mailman.18304.1211481047.23553.genealogy-dna@rootsweb.com><164510.5818.qm@web31508.mail.mud.yahoo.com><BAY106-W35038476C75DC58ED98942BCC60@phx.gbl> <ea3bd9560805221500m6328060ew978f9561d81765f@mail.gmail.com>
In-Reply-To: <ea3bd9560805221500m6328060ew978f9561d81765f@mail.gmail.com>


Here are pointers to the recent maps we have so that people can examine the data directly:

This is the U106+ project map. This is the only map of the four mentioned here that shows the reported UK datapoints. The other three maps were built specifically to look at continental datapoints.

http://www.familytreedna.com/public/U106&fixed_columns=on

This is a map I put together showing confirmed U106+ and S29 data. The yellow points indicate people who reported a country but not a location and these were usually singletons. It contains more points than the U106+ project maps because I wasn't restricted to reporting only on people who actually joined just the one project, I could sweep a number of projects looking for U106+ matches.

http://tinyurl.com/5h9zf8

This is a variant of that map with more "country-only confirmed U106/S29" datapoints to show additional concentrations from tested samples. Again, the placement of yellow balloons is not an actual location indicator, it's a "from this modern country" indicator.

http://tinyurl.com/64p74q

This was my original map where I was taking confirmed U106+/S29 and adding DYS 492=13 samples to it to see if there was any obvious geographic correlation. DYS 492=13 was proposed to me as a marker value that often identified what turned out to be U106+ samples after testing. In the U106+ project, there have been maybe 105 samples with DYS 492 measured. 2-3 have 12 repeats for this marker, 4-5 have 14, and all the rest are 13. So there may be a relationship. We had so few U106+ points on the European continent when I became interested in this that I was trying to see if adding 492=13 points was helpful. On this map, I also added in the Myres et al frequences reported per country and two frequencies reported via David Faux for Friesland and Norway. So this map is the best one to examine in terms of David's comments below, which are accurate (Austria had 22 total samples, 6 were R1b1c, and 5 of those were reported as U106+). However, this map is also what demonstrates !
the point I was making - that much of the "northern origin" claim is based on sample sizes that aren't much of an improvement on the Austrian one. I'm not sure where the cutoff should be drawn - do we ignore reports for any country with fewer than 15 U106s? 10? 5? The decision where to do the cutoff influences the perception of the resulting geography enormously.

http://tinyurl.com/4edcp4


I agree with David 100% on sample sizes and my original point was that the data presented here doesn't seem to be extensive enough to me to support any large generalizations yet, including the "northern origin" label. I would probably go for a large generalization that this clade is very widespread in Europe, actually, but that percentage measurements we have so far are too limited to show the gradients. / Tom

==================================================
> Date: Thu, 22 May 2008 15:00:54 -0700
> From:
> To:
> Subject: Re: [DNA] S21/S28 Split+m223 stuff
>
> I must point out that total R1b1c sample size was N=6 and of these who knows
> how many were relatives or from the same remote area - we know nothing of
> the sample except "Austria". It is dangerous to make large generalizations
> on very low sample sizes - unless it is of course a hypothesis which can be
> tested. It seems however that few Austrians have tested with FTDNA so we
> really need to wait to learn whether it is a true "hotspot" such as
> Switzerland is with S28. Vince and I disagree on this point.
>
> David K. Faux.
>
>
> On 5/22/08, Thomas Gull wrote:
>>
>>
>> A Myres et al reported hot spot in Austria whose percentage is actually
>> higher than that reported in Friesland. And the mapping data we have so far
>> on individuals tested U106+ actually shows more confirmed individuals in
>> southwest Germany and northern Switzerland than anywhere else except the UK.
>> This runs counter to what everyone expected to see even in the first mapped
>> datapoints, which was a concentration in northern Germany and northern
>> Netherlands. So the predictive value of the "northern hypothesis" has been
>> problematic to date.
>>
>> Any or all of these current patterns might change later as they might be
>> due to sampling bias. But the total samples for the high Norway / Friesland
>> percentage reports are actually quite small, too. Phrased another way, the
>> "northern" conclusion is largely reached by selectively deciding which data
>> to accept and which data to ignore as opposed to accepting it all. If we
>> were to go by raw data alone without assuming specific sampling biases, we'd
>> probably be looking at the UK as the S21 origin point! Everyone seems to
>> hate that possibility .
>>
>> / Tom> Date: Thu, 22 May 2008 13:55:11 -0700> From:
>> > To: > Subject: Re:
>> [DNA] S21/S28 Split+m223 stuff>>> From: Thomas Gull >>
>>>> As a side note, I think it's far from proven that S21>> originated as a
>> northern clade. Some evidence suggests it,>> other evidence contradicts
>> that.>> I suppose the evidence that supports the origin of S21+ in
>> northern Europe is that northern Europe is where S21+ is most common. What
>> is the evidence that contradicts it?>> Jim Turner>>>>>
>> -------------------------------> To unsubscribe from the list, please send
>> an email to with the word 'unsubscribe'
>> without the quotes in the subject and the body of the message
>>
>> -------------------------------
>> To unsubscribe from the list, please send an email to
>> with the word 'unsubscribe' without the
>> quotes in the subject and the body of the message
>>
>
> -------------------------------
> To unsubscribe from the list, please send an email to with the word 'unsubscribe' without the quotes in the subject and the body of the message

This thread: