PDA

View Full Version : R1b U152>L2> Z49>S8183>> Y4355



Roslav
05-28-2019, 12:28 PM
Hello to everyone on the forum!

This is my first post here and I would like to share some thoughts about Y4355 and genetic genealogy in general. Being quite interested in genetics, genealogy and history, I have taken the FTDNA 111 STR test over a year ago. My main aim was really to check for potential patrilineal cousins, optimally from the time before surnames became almost universally heritable. So by and large we are talking medieval ages, perhaps up to mid XVIIth century. My earliest known ancestor lived around 1610-1680 not too far away from Warsaw. Anyway, I got almost no matches at all. Predicted haplogruop R1B M269. Quite a let down after spending a fair sum. There were some people (24 or 25) who seemed close at the level of 12 markers, and 4 more at the level of 25 (all four not present at 12 STRs and all four at a GD of 2), but of course I have read a lot about STR convergence, which seems to be particularly frequent with R1B and quickly got to a conclusion that my 12&25 marker matches are just that. Even more so that a lot of the guys took 67 or higher tests, yet dropped out from my match list at 25 or 37 already. I knew about nevgen and checked it out, but it didn't say anything more than M269. At this point I shelved the whole thing and got back to traditional paper trail genealogy.


However, about a week ago I used Nevgen once again and to my surprise, this time I got a very specfic subclade at 97.1%, R1b U152>L2> Z49>S8183>> Y4355. Now, I have never taken the whole haplogoup hype too seriously. With people labelling various bronze age or even neolithic clades as "celtic" "viking" or "roman" on internet forums and FB, it seemed a bit of a voo-doo to me. It always seemed to me there is some fishy urge in many to impose ethnic boundaries on Y haplogroups in order to woo yet more people into more testing. Even certain FTDNA project organisers and commercial websites would use this simplified ethnic labelling (for instance shamelessly assuming most (all?) U106 in Eastern Europe comes from Mennonite/Dutch/German settlers), and of course not even 1 in 10 would mention things like pseudosegments, false positives and STR convergence. Then you get tonnes of people excited about being descendants of ancient Romans, Celts or Vikings... but how much truth is there in all those theories? And even if we discard all this ethnic banter, what is the use of knowing you likely shared a common ancestor with an Frenchmen or an Irishman, a German or a Russian all the way back to 2000BC or Proto-Indoeuropeans, or even 800BC? Bronze Age was a giant melting pot. No paper records, hundreds of potential migrations, conflicts, admixtures and so on... Literally anything could have happened. So unless we are talking about more recent times, early medieval ages or late antiquity, seemed like there is little if any advantage ;)


So I took Y4355 as an example. Seemed easy at first - only a handful of people confirmed and about 10 STR profiles publicly available. And here I run into my first difficulty - what is the actual age of the clade? The Big Tree gives an estimate of "3351.9 YBP (1403 BC). The 95% confidence interval is 2089 BC to 763 BC." Yfull gives 4100 ybp (3300 to 4900 yb). That is 2100 BC (1300-4800 BC).


In one way it does not really matter - it is still the no-good bronze age, when anything could have happened... So I gathered 10 haplotypes at 67 STRs, fed them into Nevgen TMRCA calc and it gave me an estimate of 89 generations. Roughly 3000 ybp, 1000 BC. Seemed more in line with The Big Tree estimate. So my question is - which one do you guys think is more correct? The gap between Yfull estimate deepens when you look into the branches of Y4355 (which there seem to be three). For instance the R-Y14145 branch has an estimated TMRCA of 3700 ybp. However when you feed 67 STRs from its members: Rinda/Soucek, Myers, Paul and Long into Nevgen calc it gives 60 generations (about 2000 ybp). A massive difference... I think this is of utmost importance when it comes to saying anything more specific about this, or in fact any other, clade. At the moment there seem to be 3 possibilites:


1. Y4355 as a deep rooting Neolitihic to Bronze Age relic from the Proto-Indoeuropean stage, where a single PIE ancestor left descendants in different IE branches, and through drift and bottlenecks we got to the present distribution.

2. Y4355 as a later, Bronze Age to Iron Age clade, possibly related to a particular branch of Indoeuropeans.

3. Mixture of 1&2.


When it comes to theory 2, there seems little way to reconcile the results. "Public Haplotrees" at FTDNA give 31 individuals and 25 of known origin of whom:


11 are Scandinavian
5 are German (including one Swiss)
4 are American - although some are likely of German extraction, like Myers or Reichhardt
2 are French
2 are Czech, however this is a peculiar case. The surname is Soucek/Rinda, however it turns out that ancestor's surname was originally Rinda (he inherited Soucek's house and field, and his surname too as a result): https://www.okoun.cz/boards/rodokmeny%2C_genealogie?contextId=1067528081 Rinda however is almost certainly a German surname, from "Rind" or "Rinde".
1 is Irish


Seems like one of the "unknowns" is an Italian from Friuli or Veneto.

One could speculate here about some odd Celtic migrations to Scandinavia and Poland, or slave-theories. Perhaps Germanic migrations to Ireland and North Italy, but it all seems highly random... At this stage I would say I am most inclined to take Yfull estimates and assume we are dealing with a very old clade and descendans of a very distant ancestor from the Proto-IE stage.



If Rinda/Soucek, Myers, Paul and Long are more likely to had a MRCA 2000 ybp (than 3700 ybp), then perhaps theory 2 cannot be entirely excluded. Also, I have noticed that Myers from the R-Y14145 branch and Hassel from the R-Y17276 branch (the other major branch of Y4355) seem to be most "central" out of all the ten 67 STR haplotypes when analysed by:

http://www.mymcgee.com/tools/yutility.html

Abbott-Marchbanks branch seems most off and so does Hunsperger (the Swiss guy). Hassel and Myers, at 67 STRs, have a TMRCA of ~1600 ybp (48 generations) according to Nevgen, even though they belong to different Y4355 branches and Myers' branch has 60 generations TMRCA.




Maybe someone would like to share their views on this? Are the STR-based Nevgen TMRCA estimates any credible???

MitchellSince1893
05-28-2019, 12:47 PM
In my experience, once you get beyond the 1000AD range, STR dating and matching is questionable as many STRs can back mutate to original values, thus appearing younger than they really are.

Iain McDonald comes up with this date range

Y4355 1735 BC (2517 BC — 983 BC)
http://www.jb.man.ac.uk/~mcdonald/genetics/temp/table.html


The Big Tree gives an estimate of "3351.9 YBP (1403 BC). The 95% confidence interval is 2089 BC to 763 BC." Yfull gives 4100 ybp (3300 to 4900 yb). That is 2100 BC (1300-4800 BC)

Taken together, all 3 SNP methods overlap 2089 BC to 1300 BC...or 1750 BC +/- 450 years.

Roslav
05-28-2019, 01:26 PM
Cool! Thanks for a clear answer! So it pretty much seems that not only does Y4355, but also all of its major branches (coloured very light brown below) stem from Bronze Age:

https://www.ytree.net/DisplayTree.php?blockID=1222

Then, the younger subbranches (light brown) are more recent. Like Scandinavian R-Y16875 and Rinda/Soucek's R-FGC20780. As for STR dating, seems like basically anything that falls off the FTDNA GD chart:

https://i.stack.imgur.com/M2BuH.jpg

Could be spurious and coincidental, especially when without additional SNP testing.

MitchellSince1893
05-28-2019, 02:27 PM
As an example my father has one 111 marker match at a genetic distance of 10. Another match has a genetic distance of 11. The later doesn’t show up as a FTDNA match but he a closer match to the former.

From memory, SNP dating and STR dating overlap in 1000 to 1300 AD range. The SNP dating range is 700 to 1300 AD range, while the STR method ranges from 1000 - 1400 AD. I believe our TMRCA lived in the late 1100s AD.

Roslav
05-28-2019, 06:15 PM
As an example my father has one 111 marker match at a genetic distance of 10. Another match has a genetic distance of 11. The later doesn’t show up as a FTDNA match but he a closer match to the former.

From memory, SNP dating and STR dating overlap in 1000 to 1300 AD range. The SNP dating range is 700 to 1300 AD range, while the STR method ranges from 1000 - 1400 AD. I believe our TMRCA lived in the late 1100s AD.

Interesting example! So here actually the STR estimate seemed older for the more related cousin, while made the other, more distant guy, seem closer by chance. However, I assume both of them are still relatively close matches?

All in all, sounds like without Big-Y and more samples, there is little way forward in terms of finding anything more definite. 10 or 11 in 111 looks pretty close though? I am a bit surprised the TMRCA estimate is over 1000k! Myers, Long and Paul got GDs of 12, 15 and 17 at 111 STRs (they fall pretty close to 50% of those values at 67 STR: 6, 7 and 10) and yet their TMRCA is about 1300 ybp, 1150 according to yfull. Then we got Argillander and Hassel whose TMRCA is about 1150 ybp (given at Yfull) or perhaps even 400 ybp (looks like McDonald's estimate for a whole bunch of mutations common to both of them, after: https://www.ytree.net/BlockInfo.php ) and yet they've got a GD of 13 at 67 STRs (so it would not be overly surprising to see a GD of 20 at 111, although this GD has likely been overestimated since Hassel had one funky DYS464 mutation that counts as 3+ GD; went from 15 15 17 17 to 15 15 16 16 16 17 17; if we count those multi-copy STR marker changes as single changes then it would be more like GD of 8 or 10 at 67 STRs). Anyway, seems like STR GDs can fall into a pretty wide range regardless of the real relatedness through SNPs, without necessarily either precluding or confirming any close kinship at first glance.

MitchellSince1893
05-29-2019, 01:47 AM
I'm at home now and have access to my info...as opposed to trying to remember the date ranges.

STR dating methods for my current terminal SNP branch (FGC12384) range from 950 AD to 1640 AD
SNP dating methods range from 371 AD to 1400 AD

Both methods overlap 950 to 1400 AD with a mid point of 1175 AD

SAPP tool for my branch is 1050 to 1300 AD. Ironically the same midpoint of 1175 AD
http://www.jdvtools.com/SAPP?fbclid=IwAR3n_H3WLceigIxsG3AGl59BluJXj9P1UO80 bMUzBreToKyE9HhMcRvvdDk
It was developed by anthrogenica member Dave-V.

In this post Dave's chart for 111 marker stops at a genetic distance of 20. At GD=20 the 95% range is 22 to 46 generations. https://anthrogenica.com/showthread.php?15396-Y-67-Matches&p=566142&viewfull=1#post566142

At 30 years per generation a GD = 20 would be 570 to 1290 AD. This assumes that they are valid matches and not due to "STR convergence"

The above isn't very helpful for your Y4355 inquiry, but once you go back beyond 500 AD, STR dating is of little use in my opinion.

Roslav
05-29-2019, 07:13 AM
Frankly, at this point I find like 90% stuff here helpful, so big thanks for linking Dave's post. And I think I get the point, the longer the timespan, the more convergence you may see and the less reliable STR GDs become. I guess it may work the other way too and the probabilities make the potential GD span after let's say 1k or 2k years large enough to render STRs usless in terms of an reliable TMRCA estimate.


STR dating methods for my current terminal SNP branch (FGC12384) range from 950 AD to 1640 AD.... At 30 years per generation a GD = 20 would be 570 to 1290 AD. This assumes that they are valid matches and not due to "STR convergence"


Ok, so actually the TMRCA for your Dad and that distant cousin is about 800 ybp (1175 AD) with a GD of 11 at 111 STRs. At this level and in your case, the STRs pretty much worked, judging from this chart: https://i.stack.imgur.com/M2BuH.jpg A GD of 11 should take about 28 generations, and 28 times 30 (or 33) will be about 800-900 ybp. But yeah, I get it, 10 or 11 is somewhat borderline and I guess it could have potentially been wildly off. There must be a reason for FTDNA having chosen their cut off points the way they did.



SAPP tool for my branch is 1050 to 1300 AD. Ironically the same midpoint of 1175 AD
http://www.jdvtools.com/SAPP?fbclid=IwAR3n_H3WLceigIxsG3AGl59BluJXj9P1UO80 bMUzBreToKyE9HhMcRvvdDk
It was developed by anthrogenica member Dave-V.



I typed Y4355 and it produced this image with the three major branches.

30663

Is there actually any reason why R-Y16875 looks more bold?

asquecco
06-01-2019, 12:27 PM
"Public Haplotrees" at FTDNA give 31 individuals and 25 of known origin of whom:
11 are Scandinavian
5 are German (including one Swiss)
4 are American - although some are likely of German extraction, like Myers or Reichhardt
2 are French
2 are Czech, however this is a peculiar case. The surname is Soucek/Rinda, however it turns out that ancestor's surname was originally Rinda (he inherited Soucek's house and field, and his surname too as a result): https://www.okoun.cz/boards/rodokmeny%2C_genealogie?contextId=1067528081 Rinda however is almost certainly a German surname, from "Rind" or "Rinde".
1 is Irish

Seems like one of the "unknowns" is an Italian from Friuli or Veneto.


In my opinion, 23andme could be an important source of R-Y4355 results.
Just considering my 23andme account, I have 3 autosomal matches from North East Italy classified as R-Y4355.
That's why I'm using this source to map R-U152>L2>Z49>S8183>Y17176>S20782 results:
https://bit.ly/YDNA-R-S20782
In these specific cases, I think 23andme results can be a much more solid basis compared to the available NGS results.

Roslav
06-02-2019, 09:09 AM
Hey! Thanks for pointing this out, I was about to write a new post since Ive noticed there has been some mention of Y4355 as well as other, related clades in the Balkans (and S8183 in general). For instance here:

http://www.forum.hr/showthread.php?t=935701&page=189
https://forum.poreklo.rs/index.php?topic=1436.0
https://www.eupedia.com/forum/threads/33522-What-are-Bosnian-genetics-similar-to-(comparable-country-with-almost-same-genetics)
https://www.eupedia.com/forum/threads/26201-Who-were-and-are-the-Albanians-and-their-DNA/page38


I am not sure when and how, but seems like they found some Y4355 in Western Bosnia (in 3 families? 3 villages? in one of the threads on "forum.hr" they also seem to be suggesting that those 3 fellows are somehow related to another finding in Eastern Herzegovina - not sure whether by SNP or STR), although given all that we know about this clade, it can hardly be considered "Proto-Illyrian". Altogether, there definitely seems to be a distinct and fairly diverse cluster of S8183 in North-East Italy, Slovenia and Croatia (and perhaps it may also extend into Serbia/Bosnia too). It's quite evident from the map you posted. It's probably still mariginal and rare, as most R1b in the Balkans. Whether it came there with the Slovenes or had already been there throughout the Roman times, it seems to be strengthening the hypothesis of Y4355 being a very old clade. Just like S8183:

https://www.yfull.com/tree/R-S8183/

From Puerto Rico, Portugal and Ireland to Russia, Bulgaria Slovenia and Bosnia - it looks like R-S8183 and daughter clades must have dispersed pretty early on. The only puzzling thing I find is the striking relatedness of the Bulgarian guy and the Puerto Rican - 50 ybp! It must be something very recent, and the true origin obviously points to Iberia, with the 1250 TMRCA with the Portugese guy.



That's why I'm using this source to map R-U152>L2>Z49>S8183>Y17176>S20782 results:
https://bit.ly/YDNA-R-S20782
In these specific cases, I think 23andme results can be a much more solid basis compared to the available NGS results.

Right, so this is only Y17176??? This is quite in striking contrast to:

https://www.yfull.com/tree/R-S8183/

and

https://www.ytree.net/DisplayTree.php?blockID=2058

There's tonnes more samples from Italy, Croatia and Germany, which are not even on Yfull!

What do you think about your own relatedness to Hartwicks from Canada? There again seems to be a bit of a gap between McDonald's/BigTree TMRCA estimate (around 1100 BC) and Yfull (approx 250 AD).


EDIT: ok, I thought the Russian guy is in your subclade, but actually he is in another clade (4353), while you got a TMRCA at 250 AD with a Slovene (flags got me confused), which makes perfect sense.

asquecco
06-02-2019, 10:08 AM
Hi,
I'm in touch with all the kits' owners listed in this YFull tree:
https://www.yfull.com/tree/R-Y17176/
One of the interesting facts about collecting these 23andme R-S20782 results, is to find surnames with a similar origin, like Hartwick/Hardwick or Akard/Eggert and also to be in touch with other people searching for their paternal roots, just like I'm trying to do now.
That's why I also created a Facebook group as our private "R-S20782 club", to share info and to find possible connections:
https://www.facebook.com/groups/S20782/
By the way, together with other people included in my 23andme R-S20782 map, among its members there are also all the owners of the NGS kits listed in the YFull tree above.
Here is some info about Hartwick/Hardwick paternal line you mention:
https://www.genealogy.com/forum/surnames/topics/hardwick/750/
About the S8183 result from Portugal you mention (R-CTS3409*), in the YFull tree it's listed as PT-20 Azores. From Wikipedia:
"By 1490, there were 2,000 Flemings living in the islands of Terceira, Pico, Faial, São Jorge and Flores.
Because there was such a large Flemish settlement, the Azores became known as the Flemish Islands or the Isles of Flanders.
Prince Henry the Navigator was responsible for this settlement."
Another interesting personal finding is a R-S20782 result directly from my small town where my paternal line is documented since 1551 and the other one since 1471.
Something to be investigated carefully.

Roslav
06-02-2019, 10:31 AM
This is interesting. So Hartwicks are also from Germany? I mean, so is Paul and Myers. Rinda/Soucek almost certainly too. Likely Graff and Reichhart as well. This German connection seems coming up quite frequently.

I remember somebody mention that the Barry family from Ireland are likely Flemish too - I am not sure how credible this is but these guys make up like 90% of R-Y4356, sister of Y4353. But seems pretty strange they found little Y4356 and related clades (like that one from Azores) in the Netherlands and Belgium.

In principle, the distribution of Y4355 would appear very similar to the one from your map - one cluster in NE Italy and Croatia/Bosnia to the South, one in W (SW?) Germany in the middle and another one in the UK/USA. The Scandinavian Y4355 cluster being the only difference. A somewhat more "northernly" reflection of Y17176>S20782 but still sort of along a central axis, with fairly noticable absence in both Western and Eastern Europe. I guess one could argue that's a potential reflection of the clade's "heimat" - not too far away from the Alps, and around the time of Tumulus or Urnfield culture.


This is interesting. So Hartwicks are also from Germany?.

But coming back to this, with a TMRCA of 2400-1750 ybp between you and Hartwicks, and assuming it must be predating 1500 AD that sounds much more like a genuine sign of some ancient connection - Celtic or Germanic in this case. I mean, it is even more interesting to note that the next-in-line, Slovene guy is more distant to you than the German Hartwicks.

Friuli and Wurttemberg round 400 BC to 250 AD, there can be hardly any more plausible explanations than the Celtic connection? I guess one could think potentially also of Goths and Lombards too (but then this would likely point more to Scandinavia or Northern Germany than Wurttemberg). Perhaps more recent expansions of Germans from Bavaria (which once took up most of today's Austria) through the Alpine passes from 6th-7th century onwards would explain it better. Although I personally think the Celts work best as an explanation because most of all the clades are very old, the connection is ancient and finally there is a clear cluster in W Germany and NE Italy with a marked gap in-between:

https://www.google.com/maps/d/viewer?mid=1EIjrvlTdLeMloTqSs0rj-IObJgvsbMGV&ll=47.450986529749045%2C3.0072753906249545&z=5


One way or another there is a genuine connection somewhere there, and given the similar distributions of Y4355 and Y17176 in this region (West to SW Germany & NE Italy) it is tempting to see a single factor behind their dispersal.

asquecco
06-02-2019, 12:43 PM
There are indeed different options on the table and many biases in the data to be considered.
More than the distant origin of the group, I would focus on the recent genealogy.
In my group I see some people using this map to rule out some self-reported origin of their paternal lines. Eg: is it from Normandy or from Belgium? Looking at the map I would say the latter. In my case I would say that I'm at least in the right place. I think it's a good and practical way to use it.
But I think that such kind of approach is possible only for few SNPs with not so many available results, like S20782 now.
23andme started testing it from late 2017 and moreover it seems to be a quite rare SNP.
I do not think this approach would give a great result if applied to Y4355, but I could be wrong.

Roslav
06-02-2019, 05:51 PM
There are indeed different options on the table and many biases in the data to be considered.
More than the distant origin of the group, I would focus on the recent genealogy.
In my group I see some people using this map to rule out some self-reported origin of their paternal lines. Eg: is it from Normandy or from Belgium? Looking at the map I would say the latter. In my case I would say that I'm at least in the right place. I think it's a good and practical way to use it.
But I think that such kind of approach is possible only for few SNPs with not so many available results, like S20782 now.
23andme started testing it from late 2017 and moreover it seems to be a quite rare SNP.
I do not think this approach would give a great result if applied to Y4355, but I could be wrong.

I think that is very correct! One can immediately think of three such comparisons for Y4355, that is Argillander vs Hassel, then Myers&Paul vs Rinda, vs Long. Given the history of the Swedish settlement in Finland or German in Bohemia, the first two don't present any major mystery, although Rinda's may have as well been there for 1-2k years before medieval German colonisation, given its age. The relatedness between Long from Dublin on one side and Paul&Myers from Western Germany on the other, is also interesting. One would expect him to be closer to Marchbanks and Abbott, who form a third major (and seemingly Anglo-Saxon) branch of Y4355. Instead, he falls into what appears to be an early medieval German cluster of Y4355. I very much wonder what the age of Y16875, shared by Heim&Richhart and Hassel&Argillander, may be.

Robbiem1
06-09-2019, 01:43 AM
Greetings, my name is Rob Myers, BY129625 a subclade of Y4355 (shared with Long and Paul).
I have read your posts and the discussion with much interest. But I am hopelessly lost.
Here is some info which you might find useful. I believe I am Palatine German pre ca 1750. I have not been able to make a positive connection to a Myers (Mayer etc) family yet, although I do have one family in the frame, from a little East of Karlsruhe.
I have compared my SNP details with Paul as found on Ytree (Paul not yet finalised), and used the YFull methodology to calculate TMRCA which comes close to 1250ybd, ie about 770AD. I think there may be an anomaly in the Long results which gives a distorted TMRCA.
I think Long might also be Palatine German, and quite possibly his family came out of Germany ca 1709 and ended up in Ireland (about 3300 Palatines did end up there esp. around Limerick, after first entering Ireland at Dublin). The names Myers Long and Paul all appear in shipping records of the time. So representatives of the families may have travelled at about the same time even if the actual antecedents of we three are a little different.
I am particularly interested in the use of STR's to estimate age, but I have yet to see a methodology worked through. I was under the impression that SNP were great for distant lineage, but STR's might be more useful for nearer term connections, as these mutations occur more frequently.
Would welcome any comment, helpful or otherwise to help me understand the meaning of your thread.
Incidentally, there is a fourth person (of known historical German extraction) who seems to fit with us very well, on an STR basis but he has yet to do any BigY or SNP testing.
Rob

Roslav
06-09-2019, 06:40 PM
Hey! Nice to meet you. To be honest when I started this thread, I simply wanted to get to know more about Y4355 and genetic genealogy in general (of which I have relatively little experience), and there wasn't really any superspecific aim on my mind aside from sharing some general thoughts and asking a a couple of questions, specifically about timing. Having re-discovered nevgen I got a bit excited, but could not really understand the reason behind all the different age&TMRCA estimates, like on Yfull and The Big Tree (now I understand both are SNP based) or the ones given by nevgen or McGee Y-Utility. The gap between those estimates seemed puzzling to me, and since dating is quite central to the problem of relatedness I thought someone could share their views and help me out. And perhaps share some other thoughts on Y4355. There has been no other meaning to this thread than that I guess, although then the discussion took a slightly random turn and got perhaps a bit chaotic (but then, at least for me, it's nice to have a little chat on a subject of interest even if not too specific ;))

Robbiem1
06-09-2019, 10:18 PM
Hi again, Thanks. You seem to have a far deeper understanding of that period of history and the implications of genetic connections than I ever will.
The question remains as to where my lot were about 770AD and what happened between then and say 1700AD? What were their tribal affiliations etc, what forms of governance did they live under, what was their manner of subsistence, religious practices and so on. Any thoughts?
Rob

Roslav
06-10-2019, 07:21 PM
Hi again, Thanks. You seem to have a far deeper understanding of that period of history and the implications of genetic connections than I ever will.
The question remains as to where my lot were about 770AD and what happened between then and say 1700AD? What were their tribal affiliations etc, what forms of governance did they live under, what was their manner of subsistence, religious practices and so on. Any thoughts?
Rob

Yeah, I mean this is what drew me to the subject in the first place. First time I came across a discussion about haplogroups was actually on a forum dedicated to history&archeology, some 9 or 10 years ago. But then I quickly got discouraged since there was always a lot of ethnic-origin speculations going on in those discussions, which, given the very old age of major haplogroups, seemed futile most of the time. Now with Y4355 I got a bit more personal motivation to re-visit the subject.

When I took theSTR results for 10 people and looked at them, at the level of 67 markers using McGee's tool, I noticed that for some reason You Rob seem to be the closest one to most other guys, this works for your branch when you do pairwise comparisons with Paul and Long (those two are closer to you than to each other), but also holds when you compare to Rinda/Soucek, and even Hassel who is from an entirely different branch of Y4355. Finally this also works for me ;) since with a GD of 13 at 67 STRs, you are closest in the whole set. So I assume it may have something to do with you being closest to the modal haplotype perhaps? At least this is what McGee's tool suggested here:

30885

This is 67 STRs. I have little clue whether being close to modal haplotype may mean your "line" has, by chance, not mutated as fast as others (or reversed more frequently) and if this has any bearing on potential "population of origin". I also noticed that The Big Tree gave something like this below:

"Haplotype Progression - The data below reveals the progression of changes in the haplotype. The bottom row is the haplotype for this man and above him are the inferred ancestral haplotypes for various upstream blocks."

However only for the Hassel-Ivarson-Argillander branch (when I click on Marchbanks, you or Soucek - I don't get that). Anyway, if you include this "ancestral" haplotype in the equation, results are similar, but this haplotype is way closer to modal than any other and you and Argillander are both next-in-line:

30886


I guess the only solid conclusion from this is that there are two clusters Paul-Myers-Long and Marchbanks-Abbott. The rest is less clear. For instance The Big Tree dates mutations common to you as well as Rinda/Soucek to 1072 BC. This is of course way more than any STR estimate. But as for the three of you - it would seem to be some Istvaeonic/Franconian cluster?

Roslav
06-10-2019, 07:36 PM
posted by mistake sorry!

Robbiem1
06-11-2019, 02:36 AM
Thanks for that post.
That is very interesting. I would like to share with you the results of a little exercise I did pre receiving my Big Y500 results.
Using my own y67 and y111 tests, as a baseline, I competed a regression analysis of Rinda, Long, Paul, Argillander, Marchbanks and several others under Y4355. This analysis takes into account the number of STR differences, the size of the differences, but not their location on the various markers, so its a bit 'back of the envelope' in style.
This resulted in my best fit lying between Rinda and Long, so I was not entirely surprised when the results came through that I was actually slotted in with Long. He is regarded as a match with me by FTDNA at BigY500. Mr Paul by this analysis was actually a closer fit to me than Long and remains so.

Moving on, I am aware Mr Paul has recently completed his BigY700 test. He is a match with me at Y67 together with Long, and I have no matches at Y111. He is not regarded as a match with me at BigY. Mr Long is.
However Mr Long is closer to Mr Paul and matches at Y111, but not BigY.(I am presently doing an upgrade to Y700 so that we have an apples for apples comparison).
In fact Mr Paul has no matches at BigY.
So yes, I seem to fit with everyone but match only with Long.
Now this raises an important question. and that is...what is the definition of a match?
Mr Long is shown as having 29 SNP variants difference to me. Mr Paul has some 35.
I would have guessed that with a long SNP trail leading all the way down to BY129625 that we three are all 'matches', but there is still a time gap to resolve how/when these differences arose. Rather this than some arbitrary definition that says with less than 30 variants you have a match, but with more than 30, you do not.
With much variability in the rate of formation of SNP's (as evidenced by very wide confidence limits in TMRCA throughout the entire genome analysis), it would not surprise me that such differences at this specific level could be just part and parcel of natural evolutionary factors.

Views?
Rob

Roslav
06-11-2019, 10:23 AM
Views?
Rob

Hey! Let me reply in full later, but for now I thought Id share this little finding with you (by the way Ive discovered a very decent surname mapping tool, maybe better than the old "verwandt" by MyHeritage which they took down)... So basically we know that You and Paul come from Western Germany somwhere in "Rhineland-Palatinate". Now I took some of the other surnames in the cluster, plugged them into https://geogen.stoepel.net/ and look what I got.

Rinde, Rindte, Rind, Rindt - if you cut out major cities like Berlin or Munchen from the equation than it's mostly Nordrhein-Westfalen. The first two variants are almost exclusively limited to this province, bordering on Hesse.

Fahl/Fahle is a striking example - also definitely a Nordrhein-Westfalen surname.

Graaf quite striking too - clearly Nordrhein-Westfalen, Rhineland-Pfalz, Graff also, but a bit more popular in other places.

Heim - a popular surname, but very clearly more popular in the West: Nordrhein-Westfalen, Rhineland-Pfalz, Baden-Wurttemberg and Bayern.

Reichherdt and Hartwig like Heim, although even more popular I guess...

Hunsperger - there is none, closest seems "Hinsberger" and is limited to Saarland. In this respect, it would be tempting to see whether the two French guys come from Alsace-Lorraine...

Funnily enough when you type in "Hassel" - even though they guy is apparently from Sweden and the surname is popular and I guess can be perfectly Swedish - you also get a massive blob o peaks in... right you've guessed it! No other place than Nordrhein-Wesfalen!

"Argillander" on the other hand does not sound too Finnish nor Swedish either - apparently that guy's ancestor was born in a village owned by the "von Wright" family: "Ferdinand von Wright was born at the village of Haminalahti in Kuopio, Finland. His ancestors included Scottish merchants who had settled in Narva during the 17th-Century. His father Henrik Magnus von Wright was a retired Major who owned the family estate, Haminalahden."...
This of course does not mean that the Scandinavians couldnt have inherited their own branch very early on, but those two guys seem to be stemming out a bit.
Now, Ive seen more surnames and I thought I did access them from here: https://www.familytreedna.com/public/R1b-U152?iframe=ymap but for some reason now it says "name not disclosed" everywhere. But I remember there was this guy from Schonsee near Danzig - Kabitzky - which is almost certainly a germanised version of Polish "Kapicki" "Kopicki" (a surname still popular round Gdansk/Danzig)... he must be that other guy in your cluster you mentioned (along with Rinda/Soucek)?
All in all, those surnames above still account for a significant proportion Y4355. For now it seems like this clade was inherited to a large extent by Germanic tribes, possibly at least in part through Celts. It may well be that some branches represent old bronze age lines surviving independently in relatively distant populations/tribes, however given all these facts it is tempting to see West Germany along the Rhine, namely Nordrhein-Westfalen and Rheinland-Pfalz, as a particular stronghold of this clade and a likely source of its dispersal – whether 2000BC, 1000BC or later... Perhaps it is not a coincidence that the only guy who seems to be lacking mutations beyond Y4355 has a surname (Fahle) by all likelihood from Nordrhein-Westfalen.

Roy Paul
06-11-2019, 11:24 PM
Hi This is Paul Kit# 537140 R-BY129625 it is my understanding that Y4355 evolved between the upper middle Rhine and the Danube Rivers and seems to have been part of the Bell Beaker - Hallsatt culture and Continental Celt's in southern Germany to the Swiss Alps region and up the Danube which explains our friends Rinda & Soucek under FGC20775 and there is a possibility the progenitor of Y4355 my have came from the region of what is now the Czech Republic before migrating west ? As to the GD chart Roslav provided if I'm reading it correct Rob is my only related match and vice versa also Rob is probably related to Long ! It will be interesting to see the results when Rob's BigY 700 upgrade is complete, will it bring us more recent in time or to abt 770AD as Rob has suggested ?

Roslav
06-13-2019, 04:16 PM
Hi This is Paul Kit# 537140 R-BY129625 it is my understanding that Y4355 evolved between the upper middle Rhine and the Danube Rivers and seems to have been part of the Bell Beaker - Hallsatt culture and Continental Celt's in southern Germany to the Swiss Alps region and up the Danube which explains our friends Rinda & Soucek under FGC20775 and there is a possibility the progenitor of Y4355 my have came from the region of what is now the Czech Republic before migrating west ? As to the GD chart Roslav provided if I'm reading it correct Rob is my only related match and vice versa also Rob is probably related to Long ! It will be interesting to see the results when Rob's BigY 700 upgrade is complete, will it bring us more recent in time or to abt 770AD as Rob has suggested ?

Hey Roy! Good to see you here! I messaged you on FB. I'll have more time to reply tomorrow, but I think this is an interesting case. This McGee's tool has some arbitrary thresholds. It shows two charts - one with GDs and another one which essentially the same but with GDs translated into TMRCA. And for some odd reason while it doesn't mark your relatedness to Long in colour in the first chart, it still shows you are related to both Myers and Long in the other (it actually gives values for everyone, but emphasises the closest pairs in different colours). The only difference being your TMRCA with Long is above 1000 ypb, while Long's and yours TMRCAs with Myers are estimated to be less than 1000. Anyway, the three of you form a "cluster" in the other chart with TMRCAs rather than GDs. But like the other guys here wisely pointed out, this is not a very precise estimate since it is based solely on STRs not SNPs. So it's rather a rough indicator not to be taken at face value. I guess it happens quite often that people appearing close on STRs turn out to be more related to others, with more STR differences, than to each other after SNP testing, and the other way round. Im not sure how BigY700 works to be honest. Im considering this upgrade since theyve done a discount recently.

Just from my simple perspective, I definitely think that South-West Germany is a very good if not the best candidate for the place of origin of y4355. However, I really don't know enough about Protoindoeuropean differentiation. It apparently may have been complete as late as 700-500 BC. I just googled up some info on neolithic/bronze age cultures and the variant that you described definitely makes sense. Although with dating going possibly all the way back to 2000 BC and maybe beyond, it is possible that this clade was born into a population which was neither Proto-Germanic nor Proto-Celtic. This is still a bit vague, however without going into much speculation one thing's clear - today it is predominantly a Germanic clade and a relatively small one. It may not have always been so, but I find its abscence in the West and East of some significance. It doesn't seem to be a "pan-Celtic" or a "pan-Germanic" clade. It's somewhat patchy in distribution. Perhaps that is a sign of it being not too old, or maybe it has not been very "successful"? I guess it could have dispersed multiple times, for instance very early on into Scandinavia (like 1500 BC), and then again, with the bearers of Halstatt and Latene cultures, into Northern Italy, Bohemia and perhaps also other regions in Eastern Europe. For now there are clearly two major branches (no clue how old the third, Marchbanks&Abbott branch may be), a Scandinavian one and a continental, predominantly German one, both of whom are old and predate Latene, maybe even Halstatt (if judging by Yfull estimates). I think a sensible argument in favour of this continental branch's association to Celts (rather than Franks or Istveonic Germanic tribes) is the presence of Y4355 in Northern Italy/Istria, perhaps also Western Balkans. Rinda/Soucek on the other hand is a less clear example and a bit ambiguous. If that guys original surname was Rinda than his ancestors could have as well settled in Bohemia much more recently, for instance during Medieval Ages and German colonisation.

Robbiem1
06-13-2019, 10:43 PM
Our thinking seems to be converging.
Great understanding going on here.
Rob

Roy Paul
06-14-2019, 08:16 PM
Hi Guy's & thank's for all the info, interesting theories ! with my ancestral village being Steinfeld Palatinate on the left bank of the Rhine not far from the Vosges Mountains & I believe Rob has mentioned his village could be in the Baden Wurttemberg area which makes your synopsis Roslav quite possible & that rather then Y4355 migrating west the branch FGC20775 migrated east to Bohemia. And if correct that Y4355 place of origin is southern Germany it would make some sense that when Caesar's Romans passed through the Alps to the Swiss plateau and encountered indigenous Germanic tribes they referred to them as the Keltoi. The name Κελτοί (Keltoi) and Celtae is used in Greek and Latin, respectively, as the name of a people of the La Tène horizon in the region of the upper Rhine and Danube during the 6th to 1st centuries BC in Greco-Roman ethnography.

Roslav
06-15-2019, 11:12 AM
Views?
Rob

Ok! Back to the subject!


Using my own y67 and y111 tests, as a baseline, I competed a regression analysis of Rinda, Long, Paul, Argillander, Marchbanks and several others under Y4355. This analysis takes into account the number of STR differences, the size of the differences, but not their
location on the various markers, so its a bit 'back of the envelope' in style. This resulted in my best fit lying between Rinda and Long, so I was not entirely surprised when the results came through that I was actually slotted in with Long. He is regarded as a match with me by FTDNA at BigY500. Mr Paul by this analysis was actually a closer fit to me than Long and remains so. Moving on, I am aware Mr Paul has recently completed his BigY700 test. He is a match with me at Y67 together with Long, and I have no matches at Y111. He is not regarded as a match with me at BigY. Mr Long is.


Right, so that sounds like something similar to what McGee's tool does. Is the location important though? Flicking through some forums I came accross a few posts that would mention for instance a particular, rare STR value combination for a specific set of markers, or others that would mention some markers being more stable (at least within a particular lineage). However I could not really find a good reason for this nor a list of "stable markers".
So if I understood this correctly, Paul is closer to you by this STR analysis, but in fact Long is closer - as evidenced by BigY. I actually wonder how does BigY work, I mean do you get to see a whole section of the tree and all people from your clade, for instance Y4355 or is it just your closest match?


However Mr Long is closer to Mr Paul and matches at Y111, but not BigY.(I am presently doing an upgrade to Y700 so that we have an apples for apples comparison).
In fact Mr Paul has no matches at BigY.

Ok, so there must be a threshold of SNP differences beyond which BigY feeds back no matches. So you actually cannot see a whole section of the tree nor even, when that threshold is not met, the next most related guy, no matter how many differences set you apart?

So it seems Paul is closer to you by STR, Long is closer to Paul by STR, but it's all different when you take BigY and actually You and Long seem to be more related. Interesting! So definitely Long must be of German extraction - I guess it is more likely that one German settler came to Ireland during Medieval Ages than two Irishmen to Rhineland in Germany.




So yes, I seem to fit with everyone but match only with Long.
Now this raises an important question. and that is...what is the definition of a match?
Mr Long is shown as having 29 SNP variants difference to me. Mr Paul has some 35.
I would have guessed that with a long SNP trail leading all the way down to BY129625 that we three are all 'matches', but there is still a time gap to resolve how/when these differences arose. Rather this than some arbitrary definition that says with less than 30 variants you have a match, but with more than 30, you do not.
With much variability in the rate of formation of SNP's (as evidenced by very wide confidence limits in TMRCA throughout the entire genome analysis), it would not surprise me that such differences at this specific level could be just part and parcel of natural evolutionary factors.
Views?


Yeah, from what you write it seems that this could be a very close match. I mean we are talking about an ancestor who lived about 1000 years ago. I am not sure if I get this right, but it seems to me that hypothetically two brothers could have started their own lineages and retained close STR values, however by chance one accumulated more SNP mutations - because of the high variability in mutation rate that you mentioned. I this case by chance SNPs would indicate that one of the lineages was closer to say uncle's or grandpa's brother's. Of course, it could have been the other way round - SNPs show the correct relatedness, however by chance STRs mutated back and forth more, making two lineages seem closer than they actually are... I wonder what are the odds of either STRs or BigY being more correct.

All in all, it does not matter that much, we are still talking about some really close relation. The whole trio has a TMRCA of 1000 or 1300 according to SNP analysis. I can imagine that it can be a bit less for you and Long.
So actually how did you become aware that Paul is also close, if not from BigY?



By they way, looking at this:

https://www.ytree.net/DisplayTree.php?blockID=1222

And holding to the theory that highest haplogroup diversity (and frequency possibly too?) should be present in the place of origin, then West German samples seem to fit best: Fahle (an independent lineage), Heim (the Scandinavian lineage), Paul&Myers (German lineage). It would be interesting to see where the Italian fits, or the French guys. Or other Scandinavians. I thought BigY lets you look up other fellows in your clade, shame it does not.



Hi Guy's & thank's for all the info, interesting theories ! with my ancestral village being Steinfeld Palatinate on the left bank of the Rhine not far from the Vosges Mountains & I believe Rob has mentioned his village could be in the Baden Wurttemberg area which makes your synopsis Roslav quite possible & that rather then Y4355 migrating west the branch FGC20775 migrated east to Bohemia. And if correct that Y4355 place of origin is southern Germany it would make some sense that when Caesar's Romans passed through the Alps to the Swiss plateau and encountered indigenous Germanic tribes they referred to them as the Keltoi. The name Κελτοί (Keltoi) and Celtae is used in Greek and Latin, respectively, as the name of a people of the La Tène horizon in the region of the upper Rhine and Danube during the 6th to 1st centuries BC in Greco-Roman ethnography.



They might have been Istvaeonic Germanic tribes or still Celts, Im not too well versed in ancient history unfortunately. I do not know when and how Celts were replaced by Germanic tribes, but just found two interesting maps depicting the situation in times of Caesar: https://www.ancient.eu/uploads/images/3687.jpg?v=1485681310 and https://upload.wikimedia.org/wikipedia/commons/4/4c/Cesare_prima_Gallia_58_a.C._jpg.jpg

However, having thought about this for a while I would not outright assume that all of Y4355 was inherited by Celts. Given all that is currently known about this clade, especially its age and its current distribution, there seem to be three possibilities for Y4355 dispersal (and any can be true for each individual branch of y4355):

1 through an unkown branch or perhaps still not fully differentiated branch of Indoeuropeans in early to middle Bronze Age
2 through Celts (Halstatt, Latene cultures?)
3 through Germanic tribes

Robbiem1
06-16-2019, 09:34 AM
In order to simplify the matter, I take the view that Myers, Long and Paul are all closely related at that point in time ca 1250 yrs plus or minus. The FT definition is that there should be fewer than 30 variants between men to state they are a 'match', but at this level the word 'match' has a relative meaning, not an absolute meaning.
At YFull, Mr Long is not regarded as an STR match, but he is the top of my nearest, followed by Rinda/Soucek and others seen on Ytree. However, Mr Long is regarded as an SNP match.
It intrigues me to think that the two types of genetic mutations merge in closeness, but then again it should not surprise me, since the fewer the mutations between people, the closer by definition they become.ie there has been less time for mutations to occur.
Thanks for your input.

Roslav
06-16-2019, 10:23 AM
No worries, thanks for replying to the thread. I think Im finally beginning to understand how BigY works. Partly thanks to this: https://dna-explained.com/2018/01/12/working-with-the-new-big-y-results-hg38/ By the way there is an interesting example here - a bunch of guys with a known common ancestor who lived in 1500s. One of them matches everyone on STRs (10/111, 6/67), however does not show up on anyones' BigY list. BigY is apparently assuming there is 1 SNP per 100 years. So its threshold reflects TMRCA of about 1500 years. This is an interesting example since the threshold has been crossed within less than 500 years already. The author also kinda implies not having the ability to look up all fellows you match on your terminal SNP is a setback. It seems you need to join a haplogroup project or try looking up stuff online for that.

Roslav
06-19-2019, 01:01 PM
Hey guys!

Ive re-done the STR analysis for all the Y4355 individuals I could find, that is 22 individuals. Knowing that some of the guys are definitely related I adopted two arbitrary thresholds: 12 GDs for 67 markers and 24 GDs for 111 STRs. Here is what I got:

31020


You can definitely see the main three branches of Y4355 (let's call them German, Scandinavian and British). The threshold is pretty high so some samples are almost equally close to two branches and there is some overlap as a consequence. Mainly Myears, Lorentzon (at 67) as well as Goransson and Andersson seem pretty close to many samples accross and that's the main source of ambiguity. Rinda shifts from the Scandinavian to German branch once you go from 67 to 111.

One guy from France with a surname of Egle seems pretty close to an American sample, while Allen and another sample from France (labelled "French") fall more in the Marchbanks-Abbott branch (however, they also seem close to Andersson).

To my surprise there is a guy who is even closer to me than Myers, and it is not Kabitzky (who seems to be the strongest outsider in the whole group). Apprently it's Qvarfordt from Sweden. It works the other way round too, although he is quite close to Myers too. Interestingly enough, this surname is apparently German and it seems to have no evident etymology in Swedish. One theory is that it's derived from the city of Querfurt. It was first noted around 1610 in Norway, and shortly after in Sweden. In one book on Swedish surnames I found the theory that it may be Dutch/Flemish. However, I searched a bit online and the only Querforths Ive found in the Netherlands apparently descend from Pfalzdorf in... Nordrhein-Westfalen ;)


Edit: Right, apparently it is not all that straightforward. Those Querfurths may have originated from Anhalt-Bernburg, while the etymology itself is unclear, it may be from the city of Querfurth, but may be some variation on Qvar(n)fod, or something like that, basically meaning "mill's bailiff" (from same root as "vogt" apparently). Maybe DNA will truly help to solve this puzzle!

Edit2: Turns out the branches have been marked on the R1b-U152 Project so the actual image should look like this (with 3 individuals unassigned):

31043

Kabitzky belongs to Rinda-Myers-Paul-Long cluster (but he is probably fairly deep-rooting since he is the strongest outlier in the whole group).

Dave-V
06-23-2019, 02:11 PM
Guys i hope you don't mind my butting in; I saw earlier in this thread a request for some STR-based TMRCA analysis. The McGee utility is very good for that and gives TMRCA estimates based on overall STR mutation rate averages; you can get estimates with a tighter error range if you search for Colin Ferguson's adaptation of the McGee utility and use the "Group" option there which applies individual STR mutation rates so the estimates are guided more by what specific markers have changed.

The SAPP tool also does individual STR rate-based TMRCA estimates so I took the liberty of running the publicly-available kit information for the Y4354 subgroups of the U152 project through the tool. I see earlier in this thread that you ran the SNPTREE tool on the SAPP website to draw the Y4354 subtree but that's just a graphic tool; the SAPP tool on that website analyzes STR results and draws the most likely branching tree for a group of kits, guided by the SNP results, so it integrates both SNP and STR information.

The tool's tree output for about 39 kits that I found (again this is publicly-displayed information, hope that's ok) is here (https://drive.google.com/open?id=1iWko9hoIphehCDWtKXLKLTqupgKt-sxm). Each blue box representing a common ancestor has a TMRCA estimate in green. These will not always match SNP-based TMRCA estimates and the real answer as to which is right depends on how closely either STR mutations or SNP mutations have conformed to statistical norms within any subtree - so basically without other evidence to support one or the other, either COULD be correct. However as others have noted, STR-based TMRCA estimates tend to fall off in accuracy after about 1500 years (so say before about 500AD).

I'm not saying that tree output is perfectly correct by the way because I just put in the STRs and basic SNP information available from the U152 project. If you want to run the tool yourselves, I created this spreadsheet (https://drive.google.com/open?id=175RlZmDvbbNh9Q1FarF0GaRUB9QKczGq) from the U152 project and used it (cut-and-paste) to create this TXT file (https://drive.google.com/open?id=1ZegkxJKCCep1cxUI5RB40nwZWvahbMgW) of just the Y4354 members. The TXT file is what you really need if you want to run it through the tool (http://www.jdvtools.com/SAPP/) yourselves and get the same output, then you can add or change the kits used in the run or otherwise modify the TXT file anyway you like.

SAPP has a learning curve to use it but you can review the Inputs page for an explanation of the TXT file sections and if you check the FAQ page you'll see links to videos and explanations. The "The Science" and "Credits" pages have links to various technical papers and sources for the algorithms.

Roslav
06-23-2019, 05:36 PM
Hey! I was just meant to PM you, but maybe it will be better to post it here, so here we go:

Like I wrote in my first post, about a month ago and after a yearly break, I looked up my Y-STRs and tested them using Nevgen. This time however, instead of the general M269 prediction I got nearly 98% Y4355. I've checked (and by the way discovered) yfull, the big tree, and various FTDNA projects, including U152, gathered 22 haplotypes and played around with TMRCA calculators. Now, I used Fluxus and run it at 67STRs so as to include most samples. From Yfull/Bigtree it seems there are three branches of Y4355. Comparing data on yfull/bigtree to U152 project I inferred the following:

1. First branch has two subbranches: (1)Rinda, (2)Myers&Long&Paul.
2. The other seems to be a bit more complex. It has Allen (same mutations as Heim in the big tree) &BY54610 (same as Richard) somewhere at the base, along with Egle closeby. However, it goes further into a what seems a Scandinavian subbranch with Lorentzen at its base, and then even further with additional SNPs to Hassel (same as Olsson)&Olson& two Argillanders, and from them yet further with Andersson&Goransson-Svahn. This can be more clearly seen in yfull, although it lacks names.
3. Finally the "French" guys, Marchbanks and Abbott constitute the third branch.

Now Fluxus gave me this:

31111


Essentially the first two are very similar - and a little bit different to the one you posted. Generally it separated the branches fairly well. The first branch (Rinda&Myers&Paul&Long) gets always separated and Marchbanks&Abott too, except in the last graph (where they get grafted onto the 2nd branch). The 2nd branch shows the biggest variability.

Edit: Actually the first graph seems more relevant, since I used most samples and also the modal haplotype given by The Big Tree (it's named here: MLPR-A). Generally I noticed that Fluxus seemed "unsure" where to put the "outliers" like me, Hunsperger or Qvarfordt. At 111 STRs (19 or 18 samples) it places me and Qvarfordt in the Rinda&Myers branch (branching off close to its base). Allen also gets shifted around a lot - either branching off together with Hunsperger (first graph) or separately (second graph; in those two graphs they seem to form an independent branch or two branches), or getting grafted next to me and Qvarfordt (third graph).



The tool's tree output for about 39 kits that I found (again this is publicly-displayed information, hope that's ok) is here (https://drive.google.com/open?id=1iWko9hoIphehCDWtKXLKLTqupgKt-sxm).


Thanks a lot for this! :) It seems that Myers&Paul&Long stay together in this version, but very interestingly Rinda splits off to link up with Kabitzky! Hunsperger gets sorted into the 2nd branch, while Qvarfordt falls into the 1st. Along with Egle - but that must be an anomaly since Egle should be in the 2nd, according to what they posted on U152 Project. This could potentially mean that the Qvarfordts might indeed be of German descent, and Rinda of Slavic after all.



Edit2:

Just to clarify where I got my info from. According to U152 Project, it goes like that:


1a*. "French" guys (ie "1.William 2.John 3.William 4.Christopher 5.Francis") - BY33898 et al
1a1. Abbott ( Y4355> BY33898 et al> BY55908 et al )
Marchbanks' I got from the Marchbanks Project (also FTDNA), it should be like Abbotts: BY33898 et al> BY55908 et al

2a*. Kabitzky FGC20800 et al>FGC20775
2a1. Rinda FGC20800 et al.>FGC20775> FGC20780 et al. (so indeed they are close!)
2b. Paul&Myers&Long FGC20800 et al> BY129625

3a*. Egle Y17276 et al.> BY33887 et al.
3a1. BY54610 (Richardt) Y17276 et al.> BY33887 et al.>BY54610 et al
3b. Allen Y17276 et al.> BY33915 et al> BY33899 et al
3c*. Lorentzon Y17276 et al.> Y16878 et al.
3c1. Argillander&Olson Y17276 et al.> Y16878 et al.> Y16875
3c1a.Olsson/Hassel Y17276 et al.> Y16878 et al.> Y16875> BY33885
3c1b(c1a1?). Goransson/Svahn & Andersson Y17276 et al.> Y16878 et al.> Y16875> BY42405


Hunsperger&Qvarfordt were unassigned.

Roslav
06-23-2019, 07:38 PM
I'm not saying that tree output is perfectly correct by the way because I just put in the STRs and basic SNP information available from the U152 project. If you want to run the tool yourselves, I created this spreadsheet (https://drive.google.com/open?id=175RlZmDvbbNh9Q1FarF0GaRUB9QKczGq) from the U152 project and used it (cut-and-paste) to create this TXT file (https://drive.google.com/open?id=1ZegkxJKCCep1cxUI5RB40nwZWvahbMgW) of just the Y4354 members. The TXT file is what you really need if you want to run it through the tool (http://www.jdvtools.com/SAPP/) yourselves and get the same output, then you can add or change the kits used in the run or otherwise modify the TXT file anyway you like.

SAPP has a learning curve to use it but you can review the Inputs page for an explanation of the TXT file sections and if you check the FAQ page you'll see links to videos and explanations. The "The Science" and "Credits" pages have links to various technical papers and sources for the algorithms.

Man, this is great! :)) Ive seen this tool before, but I had no clue you were the author! I was not sure how to use it. Now that you posted those examples it makes it way easier. Thanks!

serckesh
06-24-2019, 04:06 AM
Hello!
My name is Sergey and my ancestor Ankudin Ievlev (IN30992) with the haplogroup R-S8172.

serckesh
06-24-2019, 04:08 AM
Guys i hope you don't mind my butting in; I saw earlier in this thread a request for some STR-based TMRCA analysis. The McGee utility is very good for that and gives TMRCA estimates based on overall STR mutation rate averages; you can get estimates with a tighter error range if you search for Colin Ferguson's adaptation of the McGee utility and use the "Group" option there which applies individual STR mutation rates so the estimates are guided more by what specific markers have changed.

The SAPP tool also does individual STR rate-based TMRCA estimates so I took the liberty of running the publicly-available kit information for the Y4354 subgroups of the U152 project through the tool. I see earlier in this thread that you ran the SNPTREE tool on the SAPP website to draw the Y4354 subtree but that's just a graphic tool; the SAPP tool on that website analyzes STR results and draws the most likely branching tree for a group of kits, guided by the SNP results, so it integrates both SNP and STR information.

The tool's tree output for about 39 kits that I found (again this is publicly-displayed information, hope that's ok) is here (https://drive.google.com/open?id=1iWko9hoIphehCDWtKXLKLTqupgKt-sxm). Each blue box representing a common ancestor has a TMRCA estimate in green. These will not always match SNP-based TMRCA estimates and the real answer as to which is right depends on how closely either STR mutations or SNP mutations have conformed to statistical norms within any subtree - so basically without other evidence to support one or the other, either COULD be correct. However as others have noted, STR-based TMRCA estimates tend to fall off in accuracy after about 1500 years (so say before about 500AD).

I'm not saying that tree output is perfectly correct by the way because I just put in the STRs and basic SNP information available from the U152 project. If you want to run the tool yourselves, I created this spreadsheet (https://drive.google.com/open?id=175RlZmDvbbNh9Q1FarF0GaRUB9QKczGq) from the U152 project and used it (cut-and-paste) to create this TXT file (https://drive.google.com/open?id=1ZegkxJKCCep1cxUI5RB40nwZWvahbMgW) of just the Y4354 members. The TXT file is what you really need if you want to run it through the tool (http://www.jdvtools.com/SAPP/) yourselves and get the same output, then you can add or change the kits used in the run or otherwise modify the TXT file anyway you like.

SAPP has a learning curve to use it but you can review the Inputs page for an explanation of the TXT file sections and if you check the FAQ page you'll see links to videos and explanations. The "The Science" and "Credits" pages have links to various technical papers and sources for the algorithms.


You have a very good tree.

Could you add there:

269030 R-S8172 Johannes på Aga m 1519 Ullensvang HRD Norway
13 24 15 11 11-14 12 12 11 13 13 28 15 9-10 11 11 25 14 17 29 15-15-17 11 11 19-23 15 15 19 17 34-39 12 12 11 9 15-16 8 10 10 8 10 10 12 21-23 16 10 12 12 15 8 12 22 20 15 12 11 12 11 11 12 12 15 15 12 12 12 8 10 22 15 17 13 24 17 12 15 25 12 24 18 10 14 17 9 12 11

IN49394 Shchetkin Emelian, (1903-1975). Valuiki-Sv.Luchka Russia
13 24 15 11 11-11 12 12 11 13 13 29 18 9-10 11 11 25 14 17 29 15-16-17 11 10 19-23 16 14 18 16 36-36 12 12 11 9 15-16 8 10 10 8 10 10 12 23-23 16 10 12 12 14 8 12 22 20 13 12 11 13 11 11 12 12 12 12 12 12 12 11 11 12 12 12 8 10 20 15 20 13 24 17 12 15 25 12 23 18 10 14 19 9 12 11

331744 Johann Georg Kirbach, b. abt 1681 d.1721 Germany
13 25 15 10 11-15 12 12 10 13 13 29 17 9-9 11 11 24 15 17 29 16-16-17-19 11 11-23-23 15 15 18 17 35-37 12 12 11 9 15-15 8 10 10 8 8 8 10 22 15 18 13 24 14 12 15 24 12 23 18 10 14 17 9 12 11

326953 Sweden
13 24 15 11 11-15 12 12 10 13 13 28 16 9-9 11 11 24 15 17 30 16-16-17-19 11 11 23-23 15 15 18 10 8 8

746202
13 24 15 11 11-14 12 12 11 13 13 28 17 9-9 11 11 25 15 17 30 15-15-16-17 11 11-23-23 16 16 18 16 36-38 12 12

IN50024
Ambrosio Rodríguez b.1842
14 24 15 10 11-14 12 12 11 13 13 29 17 9-10 11 11 15 15 17 31 15-15-17 17 9

Dave-V
06-24-2019, 03:17 PM
You have a very good tree.

Could you add there:

269030 R-S8172 Johannes på Aga m 1519 Ullensvang HRD Norway
13 24 15 11 11-14 12 12 11 13 13 28 15 9-10 11 11 25 14 17 29 15-15-17 11 11 19-23 15 15 19 17 34-39 12 12 11 9 15-16 8 10 10 8 10 10 12 21-23 16 10 12 12 15 8 12 22 20 15 12 11 12 11 11 12 12 15 15 12 12 12 8 10 22 15 17 13 24 17 12 15 25 12 24 18 10 14 17 9 12 11

IN49394 Shchetkin Emelian, (1903-1975). Valuiki-Sv.Luchka Russia
13 24 15 11 11-11 12 12 11 13 13 29 18 9-10 11 11 25 14 17 29 15-16-17 11 10 19-23 16 14 18 16 36-36 12 12 11 9 15-16 8 10 10 8 10 10 12 23-23 16 10 12 12 14 8 12 22 20 13 12 11 13 11 11 12 12 12 12 12 12 12 11 11 12 12 12 8 10 20 15 20 13 24 17 12 15 25 12 23 18 10 14 19 9 12 11

331744 Johann Georg Kirbach, b. abt 1681 d.1721 Germany
13 25 15 10 11-15 12 12 10 13 13 29 17 9-9 11 11 24 15 17 29 16-16-17-19 11 11-23-23 15 15 18 17 35-37 12 12 11 9 15-15 8 10 10 8 8 8 10 22 15 18 13 24 14 12 15 24 12 23 18 10 14 17 9 12 11

326953 Sweden
13 24 15 11 11-15 12 12 10 13 13 28 16 9-9 11 11 24 15 17 30 16-16-17-19 11 11 23-23 15 15 18 10 8 8

746202
13 24 15 11 11-14 12 12 11 13 13 28 17 9-9 11 11 25 15 17 30 15-15-16-17 11 11-23-23 16 16 18 16 36-38 12 12

IN50024
Ambrosio Rodríguez b.1842
14 24 15 10 11-14 12 12 11 13 13 29 17 9-10 11 11 15 15 17 31 15-15-17 17 9

I can certainly add them - see this tree picture with these new 6 kits colored in orange (https://drive.google.com/open?id=155152zXaYzY54lyt8_bXcT5_6OGXr7Aa). I assumed they were all S8172+, if that's not correct let me know (it looks especially questionable for kit IN50024).

I was not able to use ALL the STRs you posted because they're not consistently in FTDNA order; I had to add the 4th DYS464 values in certain cases and truncate several kits when the markers were clearly not in the same order. If you can direct me to a project DNA Results page or somewhere these kits are listed I might be able to make use of more markers.

These are the kits with the values I was able to determine (click on picture to enlarge):
31133

Overall the placement of these kits is likely less certain given the smaller number of markers available to determine their branches; if they could be upgraded or if you had SNP data to help guide their placement there would be higher odds of correct placement.

Roy Paul
06-26-2019, 10:23 PM
Hi this a question for Roslav, could you tell us your FTDNA kit# & your oldest known ancestor & origin & what level of testing you have done & what groups you belong to ?

Roslav
06-27-2019, 08:39 AM
Hi this a question for Roslav, could you tell us your FTDNA kit# & your oldest known ancestor & origin & what level of testing you have done & what groups you belong to ?

Hey Roy! I only took the FTDNA's 111 STR so thus far and I have not joined any groups nor projects yet. Being mostly interested in recent genealogy I wanted to fish out potential cousins. Until I do a proper SNP test that would confirm my haplogroup with 100% certainty, I think I wont be joining any of the projects. Nevgen predicts Im 98% Y4355, but Id like to be absolutely sure of that (or in fact of being R1b). So I am considering getting BigY, however it is pricey ;)) and besides, now I know it has a 30 SNP limit. This sucks a bit. But it would at least confirm my position on the tree. Given all of this, and the results of pairwise STR comparisons, I think that if Im indeed y4355 then I will likely end up being some deep-rooting offshoot of one of the main branches, or perhaps an entirely new branch altogether. My oldest confirmed ancestor is a certain Stanislaus, a Polish peasant from Masovia in Poland (born ca 1610 and died some time between 1676 and 1688). He might have been son of Maciej (born ca. 1580), however the relative instability of surnames in those times precludes any true certainty. Anyway, quite an ordinary family I'd say, and, given the uneventful history of that region prior to 1600s, they had very likely been there around in 1400s or early 1500s. I will PM you my kit.

Roy Paul
06-27-2019, 11:27 PM
Roslav Nevgen is not a certainty as I learnt when it predicted me as DF27 @ 67 markers so I purchased the DF27 snp pack and I was negative.
If you do decide to join the FTDNA U152 group we can compare STR's and you would likely be able to determine if in fact you are Y4355 !

Roslav
06-28-2019, 06:12 AM
Roslav Nevgen is not a certainty as I learnt when it predicted me as DF27 @ 67 markers so I purchased the DF27 snp pack and I was negative.
If you do decide to join the FTDNA U152 group we can compare STR's and you would likely be able to determine if in fact you are Y4355 !

That is very possible, that's why Id rather get the proper test done first, and look at the SNPs. 98% seems pretty high, but I think if you actually run a test: me vs all other guys available from U152, there is still one guy from Cornwall who is even closer to me than Myers, and perhaps Qvardfordt too. However he belongs to a different clade. I did such a comparison few weeks ago. Maybe I will re-do it today just to be sure. I'll PM you.



Edit: So I've actually run my 111 STRs against all 1063 samples from the R1b U152 Project who also tested for 111 STRs, assuming infinite allele mutation model (here: http://www.dna.cfsna.net/HAP/Modified-yUtility.htm) and...

1. Almost all samples show differences of 30-50 vs mine (not more than 50 though)
2. Only about 25 out of 1063, that is 2,35% have a difference of less than 30; 7 out of those 25 are Y4355
3. None shows difference of 25 or less, except for Myers (24) and Qvarfordt (20)

Robbiem1
06-29-2019, 09:51 AM
Hi again, well that result is really quite amazing. it seems to me there is only one way to go now and that is BigY, which saves messing about with all the intermediate stuff and gets you to the end position.
As I remind myself, every time Mr Paul and I take a step forward, we find ourselves becoming closer despite the differences. You will never know otherwise. I would better value your contribution from a position of proximity.
Rob

Roslav
06-29-2019, 12:25 PM
Thanks Rob! I think that is exactly what I am going to do this summer. As for the test, mean difference between me and most U152 individuals was approx 36,8 at 111 STRs with an SD of approx 3,8 while none of the Y4355 individuals had more differences than 35 with most ranging between 27-33. Out of the 19 or so non-y4355 samples with less than 30 differences by far most were L2 with quite a few S8183 (and a few unassigned). 5 out of those 19 belonged to the Graves/Graeves family.

Robbiem1
07-01-2019, 06:41 AM
Hi guys,
Here is a question which might be quite interesting. I have been reading up on some of the background to the Keltoi peoples of Hallstatt and La Tene origins. One feature the ancient scribes have in common is that they note that these folks were tall, which I presume means relative to the general population in their experience. I wondered if many in this group likewise consider themselves to be taller than average? (you wont find the answer to this in the genetic tables, but it is an inheritable trait).
I will put my hand up for one.
Rob

Roslav
07-01-2019, 10:18 AM
Hey,

Not really, Id say about 6 ft, pretty average or just above the average. There are not many Y-linked traits, although Y genes are crucial to proper sexual differentiation and testicular development. In a way perhaps they may indirectly influence other traits. There has been much interest in SRY and its expression. It is said to be responsible for Sertoli and Leydig cell differentiation, and its expressed in the brain. However I think that such an influence must surely be minimal in comparison to autosomal genetic component otherwise it could have been easily picked up by pedigree analysis. So if we take Rhine valley as the "heimat" of U152 and its subclades, then in my case any "special" autosomal DNA, outside the usual pan-European background, that was accompanying Y4355 in its population of origin, must have been thoroughly rinsed down throughout centuries of relative isolation within the Eastern periphery of U152.

Roy Paul
07-01-2019, 01:41 PM
Interesting question Rob I believe the scribes Italians & Greeks were generally shorter peoples 5'5 - 5'7 as I have read was the average height of a Roman soldier, in my case I was 6'ft tall but now with age abt 5'11 my father was 6'1 another interesting trait that I would like to put this question forward is I Googled countries with the most bald men in the world & it comes back Japan #1 and Czech republic #2. I had premature male pattern baldness now mostly bald as my father & grandfather were, I wonder how many men under S8183 or Y4354 have the same balding inheritable trait ?

Robbiem1
07-01-2019, 10:05 PM
Yes, that was behind my question. Anyone 6ft or above I would regard as above average in height, and yes, it does raise the matter of whether this has a Y component as well as (presumably) autosomal.
The male side of my family has a strong history of tallness with my GG grandfather known to have been 6ft 1'', born 1833. Fairly tall for even that era.
I see Mr Paul is also about that 6ft mark (+/-).
As to MPB, yes there is some little evidence I have retained a full head!

Roslav
07-03-2019, 06:57 AM
Looking at all the males with my surname from my extended family there seems to be no rule for height, baldness and hair or eye colour. It comes and goes, skips a generation or two, or more. I have read that there is some evidence for X-linked maternal component to baldness, which makes the overall female contribution stronger. Id say this appears to be true in my family's case. If there is one fairly common and shared feature among males in my family it would rather be fairly deep-set eyes and quite big and long heads;)) which especially in the older generations were accompanied by long faces. It sounds funny and I dont really think it should be Y-linked, but it has been so.

Roslav
07-04-2019, 07:36 AM
They've added Qvarfordt today: https://www.ytree.net/DisplayTree.php?blockID=1222 Not finalised yet. Anyway, with an average of 130-150 years per SNP, I think I will be giving BigY a try.

Egica500
08-02-2019, 06:50 AM
Hi, i following with great interest this post. 23andme give me
R-y4355 haplo and i did y-37 reciently, there get matches with center european guys and nordic this particular sweden.
Serckesh , i Saw my GGfather, Ambrosio Rodríguez in your list , thanks.
In a future ill do y-700.i aprecciate any help or advice. Forget to say im Spaniard.

alchemist223
08-02-2019, 03:55 PM
Hello Egica500. Welcome to Anthrogenica! In order to learn more about your subclade and its origins, I suggest that you join the R-U152 project at FTDNA (here at https://www.familytreedna.com/groups/r-1b-u152/about/background).

Egica500
08-02-2019, 04:05 PM
Hi, alchemist223 . i did it, long time ago.Anyway thanks.

Robbiem1
08-03-2019, 09:22 AM
Hi, welcome to our little group. Looking forward to seeing you on the charts!
Robbiem1

serckesh
08-03-2019, 10:30 AM
I can certainly add them - see this tree picture with these new 6 kits colored in orange (https://drive.google.com/open?id=155152zXaYzY54lyt8_bXcT5_6OGXr7Aa). I assumed they were all S8172+, if that's not correct let me know (it looks especially questionable for kit IN50024).

I was not able to use ALL the STRs you posted because they're not consistently in FTDNA order; I had to add the 4th DYS464 values in certain cases and truncate several kits when the markers were clearly not in the same order. If you can direct me to a project DNA Results page or somewhere these kits are listed I might be able to make use of more markers.

These are the kits with the values I was able to determine (click on picture to enlarge):
31133

Overall the placement of these kits is likely less certain given the smaller number of markers available to determine their branches; if they could be upgraded or if you had SNP data to help guide their placement there would be higher odds of correct placement.

269030 Johannes på Aga m 1519 Ullensvang HRD Norway - R-S8172
IN49394 Shchetkin Emelian, (1903-1975). Valuiki-Sv.Luchka Russia - R-m269 - Big-Y700 is not ready yet, preliminary R-Y4355.
331744 Johann Georg Kirbach, b. abt 1681 d.1721 Germany -R-m269 - probably R-s8172
326953 Sweden - R-m269 - probably R-s8172
746202 - R-m269 - probably R-s8172
IN50024 Ambrosio Rodríguez b.1842 - R-Y4355, as said in previous posts

Roy Paul
08-06-2019, 11:04 PM
Thank's for all the info, great work serckesh it all looks good to me ! If you google Y4355 you will see The Big Tree R-Y4355 the site is Ytree.net where people upload there Big Y results other then that all I know of is U152 project @ FTDNA.

Roslav
08-15-2019, 08:02 PM
Meanwhile, I have done a quick comparison once again. This time, however, I looked up "X-fast", "fast" and "medium-fast" markers by McDonald and removed either X-fast and fast, or all three types and run the tool.

32468

McDonald's rates:

http://driscoll.dnagen.org/dna/McDonaldRates.htm

Robbiem1
08-15-2019, 10:27 PM
Hi Roslav,
Those little clusters keep re-appearing, don't they.
Well done.

Roslav
08-16-2019, 10:23 AM
Hi Roslav,
Those little clusters keep re-appearing, don't they.
Well done.

Yeah, indeed they are! And it is interesting to note that by eliminating only those 14-22 fastest markers (which together account for 14% to 21% of the whole set), gets you a very significant drop in GD. For instance for Me vs Qvarfordt it goes down by 9-13 (45-65%), for Marchbanks versus Abbott it goes down by 6-9 (60-90%), for Paul versus Long 8-10 (47-59% drop in GD), for Hassel versus Andersson 7-9 (50%-64%), for Myers versus Long GD goes down 8-10 (53-66%).

So 14-22% markers account for a 45-90% dropdown in GD. This definitely seems to support their status as fast-mutating STR markers.

This is what you get when you run the whole lot at 67 STR and eliminate either all X-fast and fast (so 14 markers in total), or just the top 10 (effectively you eliminate a bit less, since not all of those 14 or 10 are part of the FTDNA's 67 STR set).

32487

Now, you start seeing an overlap between the "German" and the "Scandinavian" branch. At 10 markers this is not yet so evident. Only "Olson" appears related to Rinda and Myers. Interestingly at this level, unlike previously, Rinda joins the Myers-Paul-Long cluster while Lorentzon joins the Hassel-Andersson-Svahn-Olsson-Argillander cluster, however Marchbanks vs Abbott "association" is somewhat weaker, and this time the "French" guys do not link up with Marchbanks-Abbott at all.

However all in all, when you look at the "111 STR minus 14 fastest", "111 STR minus 22 fastest", and "67 minus 10 fastest" comparisons, the picture is pretty consistent.

Robbiem1
08-16-2019, 10:26 PM
Hi again. This is fascinating stuff. If you drop out the fast markers as you have done, what time frame does that bring you to. In other words, can you predict when each cluster had their common ancestor?

Roslav
08-18-2019, 04:33 PM
Hi again. This is fascinating stuff. If you drop out the fast markers as you have done, what time frame does that bring you to. In other words, can you predict when each cluster had their common ancestor?

I'm afraid I am lacking adequate mathemathical expertise to give a more educated answer on this, however I simply see this transformation as a noise removing step. I have read that fast-moving markers are best for a finer differentiation of more closely related individuals, but not necessarily informative if you consider more distantly related samples. If we assume that almost all of our y4355 fellows are "related" beyond the ordinary genealogical timeframe (which seems to be the case), that is at least about 1000 ybp, it may not be overly detrimental to remove those markers.

Ive noticed that within this dataset, once you remove top 22 fastest markers, generally samples confirmed to be more closely related get a GD reduction of approximately x3 or higher. You can see this clearly with Marchbanks and Abbott as well as Andersson and Svahn. These guys, with a TMRCA of, by all likelihood, 500-700 ybp drop down quite rapidly to very low values, 1-3, not uncommon between first cousins.

The drop is more moderate (x2,5 - x3) with Long-Paul-Myers and Hassel-Andersson&Svahn, with approximate TMRCA of 1000-1500 ybp. And among others it otherwise remains generally within the x1,5 - x2,5 range.


Edit:

Those numbers will be slightly different now since the website has been updated two days ago. However, the rule generally holds I think. Now eliminating only 10 fastest markers drops down the GD between me and Qvarfordt by 8, to 12. Only French guys versus Abbott get a more significant reduction.

Eliminating top 24 markers has a very strong effect on some comparisons - most notably Myers versus Long, 'French' guys (they are really descendants of William Barbee from England, so they are not French - this is some mistake, I think someone assumed they came to England with William the Conqueror) as well as Andresson and Svahn. For those 3 pairs it drops the GD x4. Then there is Marchbanks and Abbott, Marchbanks and "French", and Me and Qvarfordt with a GD reduction x3.

All in all, it would seem to confirm the other estimates placing Long and Myers closer together in relation to Paul. That is GD 4 instead of 6. More interestingly at this level of 79 "slow"&"medium" markers, GD between Paul and Myers and Paul and Long is the same, that is 6. For Abbott&Marchbanks its 3, and Svahn&Adresson it's 2. So maybe Long and Myers have a TMRCA more like 500-1000 ybp, rather than 1000-1500 ybp.

Robbiem1
10-11-2019, 06:19 AM
Hi again,
You might like to check out Y4355 Block for details on another member.
Rob

Roslav
11-06-2019, 06:13 PM
Hi again,
You might like to check out Y4355 Block for details on another member.
Rob

Hey! Im still awaiting my results. Do you know if Weiss somehow changed the TMRCA estimate of your cluster? Did you get him in your BigY matches? Interestingly Shchetkin falls into the early Scandinavian cluster together with Lorentzon (2700 ybp according to yfull, but I guess it would be more like 2000 ybp at The Big Tree), with GDs between him and other cluster members ranging between 18 to 24. Wonder if he comes up as a BigY match to some of those guys...

Roy Paul
11-10-2019, 01:09 PM
Hi Weiss/Edwin Stumpf is my only BigY match we have 20 non-matching variants and 1 shared variant 565677, his Haplogroup is R-FT61128 where as my Haplogroup was R-BY129625 with Rob Myers and now has changed to R-FT61128 with Weiss. What I don't understand is how this breaks down to GD ?

Roslav
11-11-2019, 04:16 PM
Hi Weiss/Edwin Stumpf is my only BigY match we have 20 non-matching variants and 1 shared variant 565677, his Haplogroup is R-FT61128 where as my Haplogroup was R-BY129625 with Rob Myers and now has changed to R-FT61128 with Weiss. What I don't understand is how this breaks down to GD ?

Ok, so it seems like you have split the y4355 subclade? And now there is the two of you on one branch and Long&Myers on the other (and Rinda branching off very early). However, I assume all of you can relatively safely assume to be related and share a common ancestor within the last 1000-1500 years. "GDs" are I guess most often used to describe differences in STR comparisons. I have read somewhere that, and this is supposedly a very rough estimate, when you look at FTDNA's STR tests, then 1 STR GD=1 SNP, that is 100-150 years. Obviously this can't be be quite right since at 111 STR you do not exceed 17 or 18 GDs vs Long, but obviously you got to be more than 30 SNPs away (the BigY threshold) from both Myers and Long, since they do not come up as your BigY matches. Rob wrote that your TMRCA should be around 1300 ybp. I wonder whether by chance your branch has accumulated slightly more SNP mutations and just fell out of BigY's effective range. If Weiss comes up as a match for Long or Myers (or both) that would seem to be the case.

Roy Paul
11-16-2019, 10:24 AM
Greetings fellow Y4355 ! this is my question ? I have a GD of 9 @ 111 markers with Weiss/Stumpf and BigY STR differences of 5 of 558 with him also so a total of 14 of 669 STR differences. With and average of 125 years per STR 14x125 = 1750 ybp if I'm calculating this correctly or do SNP mutations change the calculation and if Myers and I are 1300 ybp wouldn't that make Stumpf and I much closer then 1750 ybp ?

Roslav
11-16-2019, 06:32 PM
Greetings fellow Y4355 !

Ha! I wish... but I will have to wait quite a few more weeks to find out. Unfortunately FTDNA is having issues with delivering BigY results within 8-10 weeks. Also, the likelihood decreased. I think something changed in the way Nevgen calculates the probabilites. Maybe the size of the database increased and altered the calculations. I still have no other candidate than y4355, but the probability of an "unsupported clade" increased to 20-25%. As I understand it, this most likely means Id end up as an isolated branch, likely with no BigY matches, somewhere below M269. Perhaps within or near y4355. All in all, having played quite a lot with different tools and fast vs slow marker sets, I think this is the most likely outcome, unless the connection to Qvarfordt is genuine, but this would be more like 2000 ybp to 3000 ybp.


this is my question ? I have a GD of 9 @ 111 markers with Weiss/Stumpf and BigY STR differences of 5 of 558 with him also so a total of 14 of 669 STR differences. With and average of 125 years per STR 14x125 = 1750 ybp if I'm calculating this correctly or do SNP mutations change the calculation and if Myers and I are 1300 ybp wouldn't that make Stumpf and I much closer then 1750 ybp ?

My bad! This stuff I came across at some old forum discussion. I think it was posted at the FTDNA forum. Basically, one user was saying that at 67 or 111 STRs, the number of GDs often corresponds to the number of BigY SNP differences. I can't verify this claim and it may be pretty spurious. Obviously this can't hold for as many as 669 STRs! 9 differences may well be within the last 500 years or so. That's about the same as level as Goransson and Svahn. And they are like 500 ybp according to the Big Tree, likely less than that. And (presumably) 1100 ybp according to yfull. Id definitely call that a fairly recent historical connection. If you and Weiss got like 5 SNP differences in-between then that would indicate a rather recent relation, maybe even less than 500. I thought once you get BigY results back, they give you some rough TMRCA estimate for your matches?

Robbiem1
11-17-2019, 09:45 AM
Hi, I think it more likely that with my BY700 test results I have found so many more SNP's, that by default, I have moved outside the range of 30 variants to be called a match with anyone.
Rob

Roslav
11-17-2019, 07:39 PM
Hi, I think it more likely that with my BY700 test results I have found so many more SNP's, that by default, I have moved outside the range of 30 variants to be called a match with anyone.
Rob

Well, that would make sense. Surely they must have altered the matching algorithm for Big Y-700. It has been 30 SNPs up until 2019. Any idea what the new threshold may be?

Also, from those different Y-STR comparisons, I have noticed that at 111 STRs Svahn and Andersson are clearly ahead in terms of the lowest average difference in GDs accross all other members of y4355. This still holds when you exclude all of their co-members on the "Y17276" branch, and becomes even more pronounced once you start eliminating some of the fast markers. These two are then followed by Argillander, Shchetkin and Myers (Ive excluded Long, Paul and Rinda when calculating average GD difference). It would thus seem that these guys are closest to the modal haplotype. I wonder whether this may hold any significance as to what the hypothetical population of origin for 4355 may have been.

Roslav
12-13-2019, 10:56 PM
Hey Guys, I got some news finally!

My results have arrived about a week ago and it turned out that Nevgen prediction was correct. I do belong to y4355! In fact, it seems like I am the very first y4355 sampled from my country. And what is more, this Qvardfordt connection proved to be genuine after all as well! At first, my FTDNA account displayed the y4355 "badge" and I ended up as a 4th independent single-person branch (alongside German Fahle, Swiss Hunsperger and Swedish Qvardfordt). With an average of 42 SNPs below y4355 across 40+ current members, it seemed to me like some very distant Bronze Age to Eneolithic connection. However, today I looked up my FTDNA account again and, to my surprise, my position on FTDNA's Block Tree has changed (aren't they supposed to post final results btw?).

BY-182698 - this is my current subclade and a new, 4th branch of y4355, which I have formed together with the descendant of Elias Qvarfordt. Block Tree indicates that we share about 9 "branch variants" below y4355, and we've got on average 27 private SNPs each. I got little clue as to any possible timing, but it is somewhat similar to Y17276 branch. So given the estimates on yfull or a simple "100 years per SNP" calculation that would roughly be like 1400-700 BC, I guess?

Roy Paul
12-13-2019, 11:20 PM
Congratulations Roslav ! Elias Qvarfordt is a match to me also @ GD 1 and BigY str differences 12 of 388 I'm surprised your not a match to me as well ?

Roslav
12-13-2019, 11:46 PM
Congratulations Roslav ! Elias Qvarfordt is a match to me also @ GD 1 and BigY str differences 12 of 388 I'm surprised your not a match to me as well ?

Thanks Roy! I thought you were a close match with Weiss. You sure you meant Qvardfordt? Frankly, I am not sure what is going on in here anymore. Last week my terminal SNP was simply "y4355". I ended up a lone branch, with 48 private variants.

I do not know for sure if that Swedish guy is Qvarfordt as FTDNA does not reveal the names of the participants unless they are a match (and I still have no BigY matches!). But judging from the diagram on The Big Tree and the data on U152 Project, this would seem to be the case. He, Hunsperger and Fahle appear as the 3 "lonely" branches. Now, last week I was a 4th one. But now Ive got a separate branch with that other guy and Ive got 40 private variants. This means he must have about 14 private variants since FTDNA says we both have 27 private SNPs on average. Seems a little bit disproportionate, but maybe SNPs can occur at very different rates?

JoeyP37
12-14-2019, 12:23 AM
There's a guy from Penza Oblast on the YTree R-Y4355, is that you?

Roslav
12-14-2019, 10:51 AM
There's a guy from Penza Oblast on the YTree R-Y4355, is that you?

No, that guy, Shchetkin, sits on the Y17276 branch. Im Polish, but have not joined any project yet. There are about 20 y4355 people out there in the public space, but about 40 on FTDNA. Most of those not participating in any project seem to come from Sweden, Norway and Finland, making the 17276 branch the biggest, counting about 25 people.


Congratulations Roslav ! Elias Qvarfordt is a match to me also @ GD 1 and BigY str differences 12 of 388 I'm surprised your not a match to me as well ?

Roy, do you have any rough estimate for TMRCA with Rob? Rob wrote that you two are 35 SNPs apart, and he is 29 SNPs apart from Long. With TMRCA of 1150 on yfull, that would be 40 years per SNP. I wonder how that compares to the various 100-150 years/SNP estimates that come up a lot on different forums. All those 40 y4355 fellows have, on average, about 42-43 SNPs below y4355. With a TMRCA of 4300 ybp on yfull, that would indicate 100 years per SNP - but if you count the difference between your average two members of y4355 then you get 4300/84 or /86, and 51 years per single SNP difference.

Then... if you take McDonald's estimate of y4355 age, which is 1735 BC (according to: http://www.jb.man.ac.uk/~mcdonald/genetics/p312/table.html ; confusingly, it is estimated at 1403 BC on https://www.ytree.net/BlockInfo.php?blockID=1222) and divide it by 42, you will get 83 years per SNP. This boils down to the big difference between the yfull and McDonald's estimates.

Robbiem1
12-19-2019, 07:08 AM
Firstly, congratulations on your result.
The picture is becoming a little confused.
Paul and another (Weiss) have formed a new subclade (FT61128) below By129625.
I am left with Long TMRCA by STR analysis ca 1050 years at BY129625 with 25 FTDNA variants or ca 40 yrs per variant (rule of thumb only)
Mr Long has not done BigY700.
Unfortunately we do not have Weiss STR values, but my guess he and Paul are ca 480yrs apart.
So if you have say 40 FTDNA variants from Quarfordt, you might be ca 1700 yrs apart.
Can I suggest you list with YTree ( although Mr Williamson has a long waiting list) and also U152 Project.
Rob

Roslav
12-19-2019, 12:21 PM
Firstly, congratulations on your result.
The picture is becoming a little confused.
Paul and another (Weiss) have formed a new subclade (FT61128) below By129625.
I am left with Long TMRCA by STR analysis ca 1050 years at BY129625 with 25 FTDNA variants or ca 40 yrs per variant (rule of thumb only)
Mr Long has not done BigY700.
Unfortunately we do not have Weiss STR values, but my guess he and Paul are ca 480yrs apart.
So if you have say 40 FTDNA variants from Quarfordt, you might be ca 1700 yrs apart.
Can I suggest you list with YTree ( although Mr Williamson has a long waiting list) and also U152 Project.
Rob

Thanks! It was quite a journey, but in the end I don't regret a singe dollar spent on this.

480 ybp that's very recent. That's almost or just within the range of modern surname formation time for a lot of European countries. I guess you must be right since Weiss and Roy are like 7 private mutations apart. Ive just searched through several public projects and noticed lines splittin mid XVIIth or mid XVIIIth century tend to have about 2-4 private SNPs each. 7 SNPs for Weiss and Roy that's about 18% total, counting from the first SNP below the Y4355 block. You and Long seem to be 40%. Me and Qvarfordt are 75% apart. 9 in common and then it goes 40 within my line vs 14 in his. 54 in total, 27 on average. When you divide 1050 by 40 or 450 by 18 - you pretty much get 25-26 years per 1% difference. Applying this to 75% youd get like 1-200 AD. If you do 40yrs*54 SNP differences youd get about 100-200 BC.

This would seem plausible in the light of McDonald's timing, but when looking at yfull and the other branch, I should be more in the range of Shchetkin so about 800 BC, I guess? We got a similar ~20/111 GD difference when compared to other branch co-members, and a similar SNP difference. It is 27 on average for Me and Qvarfordt, and about 30 on average for Shchetkin vs other guys from his branch. Although he does have 14 SNPs in common (I got 9), so %wise he is more like 70% apart. That would mean he ought to be more like 70 times 25-26 years - 200-300 AD. That's quite a difference when you look at him at yfull!

There is something definitely 'non-linear' and slightly confusing in the way yfull calculates TMRCA. It seems to me yfull calculates years per SNP differently for your branch, and for the Y17276 branch.

If we assume the rough "100 years per SNP" estimate and yfull's y4355 age of 4300 ybp. Then, for example Shchetkin gets 4300-1400 (14 SNPs in common with co-members) = 900BC, which is quite correct and similar to what he actually gets on yfull. Rinda gets 4300-500 = 3800, which also is very similar to what yfull indicates. This however does not seem to work for others. You and Long got 25 SNPs or so in common, which should take you to about 200 AD, not 950 AD. While Weiss and Roy would get to 900 AD. Definitely not 1550 AD.

It is not just different accross the branches, but his is also changes when we look within the Y17276 branch itself...

Take for instance Argillander, Olsson and the two Svahn and Goransson guys. They all share about 33 SNPs (and are on average about 25% SNPs apart). So they all should be related with a TMRCA at 1000 AD (counting 100 years per SNP), however Yfull gives them 550 AD (a decrease from 100 to 83 yrs per SNP). Svahn and Goransson have 37 or 38 SNPs in common and are on average 8 SNPs each apart, which is 17% (almost like Weiss and Roy), but their TMRCA is 900 AD. You'd expect 1500 AD. 17% seems a lower difference than between you and Long's (at 40%). I find this a little bit confusing to be honest. As if Yfull started off by giving about 100 years per SNP, but then it decreases progressively. It is like:


40-41 years per 1% SNP difference for Shchetkin
58 years per 1% SNP difference for the Y16875 guys above
65 years for Svahn and Goransson

For comparison, within your branch it is more like:

42 years per 1% SNP difference for Rinda
25-26 years per 1% SNP difference for you, Roy, Weiss and Long.

Weird! It starts off similar in both branches for the deep-rooting guys like Rinda or Shchetkin, but then it goes opposite! And seems to be giving more weight to each SNP difference in the Y17276 guys and less in your branch.

Seems like you can never be too sure what's gonna happaen unless you upload to yfull and find the answer for yourself. Taking the extreme numbers from above I can be anything from 200 AD to 2500 BC! :P Interestingly, dividing the yfull's y4355 age of 4300ybp by the average of 42 SNPs below Y4355 in 40+ members, does give you about 100 years per SNP ;)

Could any of you guys just breifly explain what are the benefits of joining the U152 project? Maybe I should go yfull too?

Robbiem1
12-19-2019, 11:32 PM
Could any of you guys just breifly explain what are the benefits of joining the U152 project? Maybe I should go yfull too?

I will try.
U152 is primarily an academic research Project and does not offer a chat function. However, you would be placed presumably alongside Qwarfordt in the tables. Your STR values would also become visible for the likes of 'Dave-V' to incorporate in the work he has done on assessing TMRCA via Y111 STR's. You might recall his chart.

YFull offers a complete interpretation service, including the ability for subscribers to interrogate their data base, calculates TMRCA, matching etc etc. Their TMRCA method is tightly defined, (it only considers SNP's in the combBED region) and seems really good for distant TMRCA's. I think where we are, with a number of us within that 1000 year mark, it might break down a bit. However the underlying fact is the closer you are genetically to some one, the more variants you have in common and the fewer variants not in common.
I think its easy to become a bit confused between the two methodologies for calculating TMRCA's (STR vs SNP's) and the usage of variants interchangeably with specific SNP counts.
Since I am not an expert, that's why I refer to rules of thumb etc.
I would also add that chart 'Dave-V' produced some months ago seems pretty good. Rob

Roslav
12-26-2019, 01:59 AM
Thanks a lot Rob, that was helpful. I have also spent a couple days doing some online reading, including a very helpful thread here on the forum called "Big Y-700", and I think I will wait for Qvarfordt to upgrade to BigY-700 first. It seems to me that he has done Big Y-500 as he is coded "BigY2" on The Big Tree.


That would well explain why he has only 23 SNPs, including 14 private, which is very low for y4355. If you take McDonald's y4355 age and divide it by 23 you get 150 years per SNP. A well-known estimate. On the contrary - when you divide it by 49 (like I've got) - you get about 70 years per SNP. The average for the whole clade (it however includes Big Y-500s too) is 42 SNPs and gives you about 83.3 years per SNP.

Funnily enough, I have made those simple calcs two days ago and only now I have read Dave-V's post here (https://anthrogenica.com/showthread.php?16412-Big-Y-700&p=612133&viewfull=1#post612133) (great discussion by the way, don't know if you read it?) - and Dave-V has done a far bigger analysis on another clade. And he also came up with an estimate of 70 years per SNP for Big Y-700. Of course it depends a lot on the clade/age you choose. However, Y-DNA Warehouse also indicates 83 years per SNP, and it seems it may sometimes go down to even 50 years per SNP or even lower (as indicated by other posters in the linked thread, and also by some examples on the web, including "dna-explained" blog). I was actually quite surprised how well those numbers - 70 and 83 - correspond to the situation you can observe within y4355 when using McDonald's timing!

With 9 SNPs in common and 14 vs 40 "private" - we may have a TMRCA anywhere between 1AD to 1000 BC (by McDonald's timing) or very roughly 1700 BC to 500 BC by yfull's. In the named thread, Dave-V wrote that Big-Y 700 can get you both new upstream, older SNPs as well as more recent, "private" ones.

The STR consensus (by both Nevgen and McGee's utility and also Dave-V's chart, which you have mentioned) would be more like 1-150 AD, but we know STRs are unreliable.

It can only get more precise once he gets 700. We share 9 variants and we may either share more or have a lot more private ones. This would perhaps help to tighten that likely TMRCA span. Shame I can't contact him and ask whether he is planning to upgrade!

Emqew
01-19-2020, 08:48 PM
Hi. I expect they will be adding another Qvarfordh soon, ie me :-). Recently got my BIG-Y results and with the help of those results YFull has, at least preliminarily, placed me in R-Y4355. Let's see when I turn up...

Roslav
01-23-2020, 09:05 PM
Hi. I expect they will be adding another Qvarfordh soon, ie me :-). Recently got my BIG-Y results and with the help of those results YFull has, at least preliminarily, placed me in R-Y4355. Let's see when I turn up...

Hello, and welcome to our thread! I noticed a new tester joined R-BY182698 on FTDNA's Block Tree a couple weeks ago, altered the whole branch and appeared to be closely related to Qvarfordt. Good to see you on Anthrogenica! it's unfortunate I have not used yfull yet, but Im considering it. It would be very interesting to see what you get. Judging by the number of shared as well as private combBED SNPs within our branch, we should end up with a tmrca likely somewhere between 500BC-1000BC (I simply used the named "shared variants" from the Block Tree and their ISOGG positions, as well as my private variants' positions to check how many of them are combBED based thanks to the spreadsheet kindly shared by Dave-V here: https://anthrogenica.com/showthread.php?13758-Spreadsheet-for-combBed-SNPs-and-Adamov-based-TMRCAs-Hg38-coordinates). This is a rough estimate. Im sure yfull will find more useful SNPs ;)

Emqew
01-29-2020, 10:53 AM
Hi. Really new at this, so any suggestions or requests are highly appreciated. Will gladly share any information you might need. My, at least currently, main objective as far as DNA research is concerned, is to try to get closer to the geographical origin of my Qvarfordt ancestors...

Roslav
01-30-2020, 02:32 PM
Hi. Really new at this, so any suggestions or requests are highly appreciated. Will gladly share any information you might need. My, at least currently, main objective as far as DNA research is concerned, is to try to get closer to the geographical origin of my Qvarfordt ancestors...

Ive been following the subject of genetic genealogy for a couple of years now, but I'm still a beginner. However, Ive spent some time on history, archeology and also linguistics. Actually the surname "Qvarfordt" is an interesting case itself! Id be glad to help with any info too, and as best as I can! I will PM you.

Roslav
01-30-2020, 05:13 PM
...duplicated by accident, sorry!

Dolomites
06-10-2020, 06:58 PM
Hello everyone! congratulations for the analysis! I am Y4355, from Friuli ... my surname was originally Mikelaç (in Friulian) .... Mikaelaz or Michelaz is also written on some documents ....

Roslav
06-21-2020, 09:42 AM
Hello everyone! congratulations for the analysis! I am Y4355, from Friuli ... my surname was originally Mikelaç (in Friulian) .... Mikaelaz or Michelaz is also written on some documents ....

Hey! That's great. First Italian on Y4355! Are you by any chance planning to do BigY at some point? It would be interesting to see which subbranch you may belong to. Do you think Mikelac may be of a Slovene/Croat origin?

asquecco
06-21-2020, 11:16 AM
Hey! That's great. First Italian on Y4355! Are you by any chance planning to do BigY at some point? It would be interesting to see which subbranch you may belong to. Do you think Mikelac may be of a Slovene/Croat origin?

I suppose there are many Y4355 from North East Italy: just among my 23andme matches I have 4 of them from this area (Dolomites included) + 1 from Austria.

Roslav
06-21-2020, 01:16 PM
I suppose there are many Y4355 from North East Italy: just among my 23andme matches I have 4 of them from this area (Dolomites included) + 1 from Austria.

That's interesting. I also read on one forum that Y4355 is likely also present among Serbs, or Bosniaks. Although Im not sure how credible this may be... Dolomites that's more like South Tirol. Shame none seems to be present on FTDNA! It would be very interesting to see if they form a separate branch, or perhaps get disperesed among other branches...

tcmyers
09-10-2020, 06:16 PM
Greetings, my name is Rob Myers, BY129625 a subclade of Y4355 (shared with Long and Paul).
I have read your posts and the discussion with much interest. But I am hopelessly lost.
Here is some info which you might find useful. I believe I am Palatine German pre ca 1750. I have not been able to make a positive connection to a Myers (Mayer etc) family yet, although I do have one family in the frame, from a little East of Karlsruhe.
I have compared my SNP details with Paul as found on Ytree (Paul not yet finalised), and used the YFull methodology to calculate TMRCA which comes close to 1250ybd, ie about 770AD. I think there may be an anomaly in the Long results which gives a distorted TMRCA.
I think Long might also be Palatine German, and quite possibly his family came out of Germany ca 1709 and ended up in Ireland (about 3300 Palatines did end up there esp. around Limerick, after first entering Ireland at Dublin). The names Myers Long and Paul all appear in shipping records of the time. So representatives of the families may have travelled at about the same time even if the actual antecedents of we three are a little different.
I am particularly interested in the use of STR's to estimate age, but I have yet to see a methodology worked through. I was under the impression that SNP were great for distant lineage, but STR's might be more useful for nearer term connections, as these mutations occur more frequently.
Would welcome any comment, helpful or otherwise to help me understand the meaning of your thread.
Incidentally, there is a fourth person (of known historical German extraction) who seems to fit with us very well, on an STR basis but he has yet to do any BigY or SNP testing.
Rob

Greetings from Thomas Myers, confirmed as a R-Y4355 yesterday. I happened upon this thread and was excited to see the discussion of another Myers with the same SNP and can share a bit about my ancestry back to Germany. My farthest known ancestor is Frederick Daniel Myers, b. 1782, immigrated to Orangeburg, South Carolina, and then on to Perry County, Mississippi via wagon train c. 1810 and finally to Mobile, Alabama where he died in 1865. My Myers lineage has been there ever since. "Daniel" as he appeared to be called, had two brothers, David Levi Myers b. 1772 and John James Myers b. 1776. Frederick Daniel's tombstone said that he was born in Hamburg, although family oral history was that his granddaughter did the inscription from her memory a good bit later. The Orangeburg connection would have them most likely coming from southwest Germany / Switzerland area, which is not far from your Palatine migration. I have not finished reading the full thread but will look forward to tracking going forward. Happy to answer any questions or provide additional info as helpful. // Tom

Robbiem1
09-10-2020, 10:40 PM
Hello Tom,
feel free to contact me directly on [email protected]
Best Rgds
Rob

Roslav
09-12-2020, 12:14 PM
I have made a new diagram of Y4355 using Dave Vance's SAPP algorithm. This is pretty much what it looks like:

39561

I guess it may be prudent to pay more attention to the generation span TMRCAs. Id say an average male generation can be about 33 years. With that assumption, it would give us a TMRCA for Y4355 at about 3234 ybp, 2740ybp to 3700 ybp. For the same set of individuals (had to manually alter Olsson's DYS464), NevGen Probability Calculator of TMRCA gives an average of 95 generations, so 3135 ybp (with the same assumption), very close. Upper estimates are within or quite close to Alex Williamson's and Iain McDonald's TMRCA estimates (3400 ybp and 3750 ybp).



Just out of pure curiosity, I also did this version - with no SNP data whatsoever:

39562

TMRCA increases to 3750 ybp on average, and it seems to be splitting the Y17276 branch, while keeping most others pretty much intact.

Roslav
09-21-2020, 05:37 PM
As I see it, this tree looks a bit like a diagram of Proto-Germanic unity, followed by splits, with German (red), Swedish (blue) and Anglo-Saxon (violet) clades emerging, presumably still some time before the Middle Ages.

39684

Especially the way R-Y17276 branch looks like, is suggestive. It is tempting to see it as illustrative of 3 subsequent linguistic splits:

1. North vs West vs East Germanic - some time in the BCs up to 150 AD?

2. German vs Anglo-Saxon - some time in the ADs, most likely 150-600 AD?

3. Swedish vs Finland Swedish - some time between 1000-1400 AD?

39685