PDA

View Full Version : L1029 in Britain?



Brent.B
11-25-2015, 01:12 AM
Hey all

I recently found out that I have a big-Y match with an Englishman (My paternal line leads to England as well). We both belong to L1029, and we apparently share 1/2 of our SNPs downstream of L1029. As far as I understand, the TMRCA of L1029 is 2100 years, and there are some that say it is 10-20% older than that (so 2500 years).

Given that we share half of our SNPs downstream of L1029, that would indicate a MCRA living around 950AD to 750AD (depending on the TMRCA used).

The goal of my genetic research is to understand how my line of L1029 came to England. Given that we have a MRCA at 950AD to 750AD, it would seem that those dates are the latest possible time for my ancestors to have entered England, and could at least theoretically have entered even before that.

Given the rarity of L1029 in Britain (and the rest of Western Europe), I am curious to see if anyone has any thoughts on how it could have arrived so early? If the 2500 TMRCA of L1029 is correct, it would even predate the first viking raids in England!

any thoughts?

rms2
11-25-2015, 01:33 AM
You would probably get more responses if you posted that in the R1a subforum.

lgmayka
11-25-2015, 03:21 AM
Given that we have a MRCA at 950AD to 750AD, it would seem that those dates are the latest possible time for my ancestors to have entered England, and could at least theoretically have entered even before that.
I would not necessarily conclude that, because an entire clan (with a TMRCA hundreds of years before) could cross into England en masse. But even if your assertion is true, it leave the obvious possibility of the Anglo-Saxon invasion.

Artmar
11-25-2015, 01:47 PM
I would not necessarily conclude that, because an entire clan (with a TMRCA hundreds of years before) could cross into England en masse. But even if your assertion is true, it leave the obvious possibility of the Anglo-Saxon invasion.
I think that it is fairly possible that somewhat less frequent L1029 sublineage got to Denmark, assimilated and then landed in England with other Danes. It could've also explained a presence of some other "Slavic" branches of R1a, like S18681 or Y2902.

rms2
11-25-2015, 01:56 PM
I don't want to be accused of attributing all R1a to Slavic peoples, but I recall that Gwynn Jones, in his excellent book, A History of the Vikings, mentioned the little known fact that there were Wends among the Anglo-Saxons. Apparently they were allies with some of the Germanic tribes in that confederation. Of course, no doubt many of the Germanics were R1a themselves. Any evidence of L1029 among the Wends or their descendants?

Brent.B
11-25-2015, 08:11 PM
I would not necessarily conclude that, because an entire clan (with a TMRCA hundreds of years before) could cross into England en masse.

Yes, I was thinking that as a possibility as well. However, I would be surprised to find a tribe only in England (where the occurrences of such DNA, if it is Slavic, should be rare), while finding none in it's homeland of Central Europe/ the Baltic. Doesn't it make more sense that a rare/early sub lineage of L1029 made it to somewhere like Denmark and expanded (slightly) once it got to England?


I think that it is fairly possible that somewhat less frequent L1029 sublineage got to Denmark, assimilated and then landed in England with other Danes. It could've also explained a presence of some other "Slavic" branches of R1a, like S18681 or Y2902.

That is my thoughts. If true, the question is when did this happen? Pre-viking? (Anglo Saxon as lgmayka said) or later?


I don't want to be accused of attributing all R1a to Slavic peoples, but I recall that Gwynn Jones, in his excellent book, A History of the Vikings, mentioned the little known fact that there were Wends among the Anglo-Saxons. Apparently they were allies with some of the Germanic tribes in that confederation. Of course, no doubt many of the Germanics were R1a themselves. Any evidence of L1029 among the Wends or their descendants?

I looked through the book but can't find reference to the Wends among the Anglo Saxons? I know they participated with the Vikings, but I have never heard that the Slavs partook in the Anglo Saxon migrations.

Michał
11-25-2015, 09:53 PM
That is my thoughts. If true, the question is when did this happen? Pre-viking? (Anglo Saxon as lgmayka said) or later?

I would wait for the TMRCA age estimate provided by YFull before trying to answer your question. Having said that, I consider the Viking/Danish variant the most likely scenario at the moment. For example, Cnut the Great is commonly considered to have been a son of a Slavic/Polish princess (daughter of Mieszko I and sister of Bolesław the Brave), so she likely was accompanied by some Poles when traveling to Denmark to marry Sweyn Forkbeard.

Brent.B
11-25-2015, 10:13 PM
I would wait for the TMRCA age estimate provided by YFull before trying to answer your question. Having said that, I consider the Viking/Danish variant the most likely scenario at the moment. For example, Cnut the Great is commonly considered to have been a son of a Slavic/Polish princess (daughter of Mieszko I and sister of Bolesław the Great), so she likely was accompanied by some Poles when traveling to Denmark to marry Sweyn Forkbeard.

Like you say, it will be more clear once we have more solid TMRCAs.

But if we assume your suggestion that Y-Full underestimates the age of L1029 by 10-20% is correct, AND the TMRCA of this English subclade is half that of L1029 (750AD), then that would predate even the first viking raids in England (they only started in the 790's)

Cnut himself was born in 990AD. This is even later than the English L1029 MCRA assuming Y-Full's age is correct with no 10-20% extra in age (950AD). So in that case, perhaps the Danelaw?

But you are right, I'll need to wait until there is a more solid picture of the age of L1029 and my subclade to make a better guess as to what this all means.

Huntergatherer1066
11-25-2015, 11:00 PM
I don't want to be accused of attributing all R1a to Slavic peoples, but I recall that Gwynn Jones, in his excellent book, A History of the Vikings, mentioned the little known fact that there were Wends among the Anglo-Saxons. Apparently they were allies with some of the Germanic tribes in that confederation. Of course, no doubt many of the Germanics were R1a themselves. Any evidence of L1029 among the Wends or their descendants?

There is this study on Sorbian genetics: http://www.nature.com/ejhg/journal/v19/n9/full/ejhg201165a.html but it looks like only autosomal DNA. It does cite a few other studies that did involve Y-DNA though that may be worth looking at but they likely all predate the discovery of L1029 and are from the STR era. Speaking of Wends, I remember driving through a historic Wendish community in Texas on a family reunion: https://en.wikipedia.org/wiki/Wends_of_Texas I wonder if any of these Texan Wends have done testing.

Baltimore1937
11-27-2015, 07:20 AM
Old "Bluetooth", Forkbeard's daddy, hired slavic mercenaries because they were loyal directly to him. He couldn't trust his scheming palace guard, etc., what with the constant palace intrigue going on.

Brent.B
12-03-2015, 07:22 PM
I don't want to be accused of attributing all R1a to Slavic peoples, but I recall that Gwynn Jones, in his excellent book, A History of the Vikings, mentioned the little known fact that there were Wends among the Anglo-Saxons. Apparently they were allies with some of the Germanic tribes in that confederation. Of course, no doubt many of the Germanics were R1a themselves. Any evidence of L1029 among the Wends or their descendants?

Your comment here lead me to do some more digging around about the Anglo-Saxon invasion, and I came across an old book called "Origin of the Anglo-Saxon race : a study of the settlement of England and the tribal origin of the Old English people". This book is OLD (1906) and uses a lot of antiquated methodology (and is worth reading with a grain of salt imo), but I did read some interesting parts that might still be a little useful.

Apparently, Bede writes that among the Anglo-Saxons, the Rugii were some of the people who came and settled in England during the Anglo-Saxon migration period. The author of the book goes on to say these Rugii (among others living on the baltic) were actually Slavs (not "Teutons" as previously thought). While this is a question that still is looking to be answered (is Przeworsk/Wielbark M458+ or not...), if aDNA does show any M458 then perhaps this information from Bede could be the first bit of evidence to show Slavic settlement in Britain?

leonardo
12-03-2015, 07:50 PM
Your comment here lead me to do some more digging around about the Anglo-Saxon invasion, and I came across an old book called "Origin of the Anglo-Saxon race : a study of the settlement of England and the tribal origin of the Old English people". This book is OLD (1906) and uses a lot of antiquated methodology (and is worth reading with a grain of salt imo), but I did read some interesting parts that might still be a little useful.

Apparently, Bede writes that among the Anglo-Saxons, the Rugii were some of the people who came and settled in England during the Anglo-Saxon migration period. The author of the book goes on to say these Rugii (among others living on the baltic) were actually Slavs (not "Teutons" as previously thought). While this is a question that still is looking to be answered (is Przeworsk/Wielbark M458+ or not...), if aDNA does show any M458 then perhaps this information from Bede could be the first bit of evidence to show Slavic settlement in Britain?

This leads to the endless debate: was L1029, via Slavs or proto-Slavs, in the area of Rugen in the 5th and/or 6th century (autochthonous)? If not, did it possibly migrate to the western Baltic by the prescribed time frame?

Brent.B
12-03-2015, 08:37 PM
This leads to the endless debate: was L1029, via Slavs or proto-Slavs, in the area of Rugen in the 5th and/or 6th century (autochthonous)? If not, did it possibly migrate to the western Baltic by the prescribed time frame?

Thankfully aDNA is starting to uncover these things. (Hopefully) soon we will have aDNA on these cultures to know for sure

lgmayka
12-04-2015, 03:11 AM
We know that R1a-M458 was on Usedom (https://en.wikipedia.org/wiki/Usedom)by 1200 A.D., according to Ancestral Journeys (http://www.ancestraljourneys.org/medievaldna.shtml).

Brent.B
12-12-2015, 03:49 AM
We know that R1a-M458 was on Usedom (https://en.wikipedia.org/wiki/Usedom)by 1200 A.D., according to Ancestral Journeys (http://www.ancestraljourneys.org/medievaldna.shtml).

Do you happen to know if settlements on Usedom continued throughout the migration period? I remember reading some places (such as Rugen) showed signs of continued settlement.

Brent.B
08-05-2016, 01:33 AM
Apparently, Bede writes that among the Anglo-Saxons, the Rugii were some of the people who came and settled in England during the Anglo-Saxon migration period. The author of the book goes on to say these Rugii (among others living on the baltic) were actually Slavs (not "Teutons" as previously thought). While this is a question that still is looking to be answered (is Przeworsk/Wielbark M458+ or not...), if aDNA does show any M458 then perhaps this information from Bede could be the first bit of evidence to show Slavic settlement in Britain?

I did some more digging around about Thomas William Shore's statement that Rugians settled in England. In the book, he says "The chief Old English names which appear to refer to them (Rugians) in Domesday Book are Ruenore in Hampshire... and Rugarthorp in Yorkshire."

I couldn't find any reference to Rugarthorp anywhere for a long time, but then I stumbled across "The Yorkshire Archaeological Journal, Volume 13". In this book, it is stated "Rogerthorp was included with badsworth and upton in domesday : But it is now part of the township of Thorp [Audlin]”.

Upton/Badsworth/Thorpe Audlin are all only 15ish miles away from Batley, where my ancestors are from. Probably a coincidence... but perhaps an interesting connection

Also, the lord of Thorpe Audlin in 1066 was "Alsi son of Karski". Sounds Slavic?

lgmayka
08-05-2016, 02:36 AM
Also, the lord of Thorpe Audlin in 1066 was "Alsi son of Karski". Sounds Slavic?
This database of Polish surnames (http://www.herby.com.pl/naz_query.html) (essentially a telephone book) shows 1554 hits for the Karski surname.

Baltimore1937
08-05-2016, 02:55 AM
I tried looking up Kollinski, allegedly in my Springer line in Medieval times (a wife from Silesia). But I only got back some words in red with 3 exclamation marks, ha ha.

JMcB
08-05-2016, 03:53 AM
Old "Bluetooth", Forkbeard's daddy, hired slavic mercenaries because they were loyal directly to him. He couldn't trust his scheming palace guard, etc., what with the constant palace intrigue going on.

That reminds me of a paper I read a while ago.

Who was in Harold Bluetooth’s army? Strontium isotope investigation of the cemetery at the Viking Age fortress at Trelleborg, Denmark.

The circular fortress of Trelleborg on Zealand in Denmark is well known as a military camp with a key role in the formation of the Danish state under Harald Bluetooth in the tenth century AD. Taking a sample of 48 burials from the fort, strontium isotope analysis once again demonstrates its ability to eavesdrop on a community: at Trelleborg, the young men in its cemetery were largely recruited from outside Denmark, perhaps from Norway or the Slavic regions. Even persons buried together proved to have different origins, and the three females sampled were all from overseas, including a wealthy woman with a silver casket. Trelleborg, home of Harald Bluetooth’s army, was a fortress of foreigners with vivid implications for the nature of his political mission.

http://www.academia.edu/3511668/Who_...leborg_Denmark

Waldemar
08-05-2016, 06:16 AM
Who was in Harold Bluetooth’s army?

http://www.anthrogenica.com/showthread.php?7096-Athanasiadis-et-al-2016-Genetic-History-of-Denmark


I tried looking up Kollinski, allegedly in my Springer line in Medieval times (a wife from Silesia). But I only got back some words in red with 3 exclamation marks, ha ha.

http://www.moikrewni.pl/mapa/kompletny/koli%25C5%2584ski.html

http://www.moikrewni.pl/mapa/kompletny/kolinski.html

lgmayka
08-05-2016, 09:13 AM
I tried looking up Kollinski, allegedly in my Springer line in Medieval times (a wife from Silesia). But I only got back some words in red with 3 exclamation marks, ha ha.
The database I cited (http://www.herby.com.pl/naz_query.html) requires exact use of Polish diacritical marks. If you can't enter those, you can use ? for a single letter and * for multiple letters.

The Koliński surname (note the accent over the n) gets 2685 hits, with the most near Warsaw, Bydgoszcz, and Gdańsk:
Wa:310, BP:7, BB:10, By:254, Ci:21, Cz:14, El:43, Gd:196, Go:46, JG:38, Kl:117, Ka:98, Ki:29, Kn:147, Ko:5, Kr:37, Ks:2, Lg:60, Lu:17, Ło:81, Łd:149, NS:4, Ol:25, Op:28, Os:2, Pl:82, Pt:1, Pł:75, Po:143, Ra:21, Rz:3, Sd:7, Sr:134, Sk:70, Sł:51, Sz:154, Tb:16, Ta:4, To:55, Wb:28, Wł:14, Wr:51, ZG:36

Baltimore1937
08-05-2016, 10:38 AM
The database I cited (http://www.herby.com.pl/naz_query.html) requires exact use of Polish diacritical marks. If you can't enter those, you can use ? for a single letter and * for multiple letters.

The Koliński surname (note the accent over the n) gets 2685 hits, with the most near Warsaw, Bydgoszcz, and Gdańsk:
Wa:310, BP:7, BB:10, By:254, Ci:21, Cz:14, El:43, Gd:196, Go:46, JG:38, Kl:117, Ka:98, Ki:29, Kn:147, Ko:5, Kr:37, Ks:2, Lg:60, Lu:17, Ło:81, Łd:149, NS:4, Ol:25, Op:28, Os:2, Pl:82, Pt:1, Pł:75, Po:143, Ra:21, Rz:3, Sd:7, Sr:134, Sk:70, Sł:51, Sz:154, Tb:16, Ta:4, To:55, Wb:28, Wł:14, Wr:51, ZG:36

I'm not very certain of that name, anyway. The trees at Ancestry don't all agree with each other. Hopefully the Springer male line is accurate. Most trees show Johanna von (or of) Kollinski in Silesia (Schlesien). She may have been the daughter of one of the Prussian barons who took over Silesia back then. But I haven't found her parents.

leonardo
08-05-2016, 10:59 AM
I found this by doing a Google search:
https://books.google.com/books?id=y7JCAAAAYAAJ&pg=PA86&lpg=PA86&dq=Rugians+slavs&source=bl&ots=pLlHegqJcF&sig=R68vaTAp13KX9_8uC9i5XKY0wC8&hl=en&sa=X&ved=0ahUKEwjpk7rXjqrOAhUJ6SYKHanLC-UQ6AEIUDAI#v=onepage&q=Rugians%20slavs&f=false.
The Rugians were considered East Germanic. However, Rugii is also associated with the rani tribe, who was Slavic. By the end of the millennium (1000 CE), the area in question would have been unquestionably Slavic.

Waldemar
08-05-2016, 12:26 PM
I found this by doing a Google search:
https://books.google.com/books?id=y7JCAAAAYAAJ&pg=PA86&lpg=PA86&dq=Rugians+slavs&source=bl&ots=pLlHegqJcF&sig=R68vaTAp13KX9_8uC9i5XKY0wC8&hl=en&sa=X&ved=0ahUKEwjpk7rXjqrOAhUJ6SYKHanLC-UQ6AEIUDAI#v=onepage&q=Rugians%20slavs&f=false.
The Rugians were considered East Germanic. However, Rugii is also associated with the rani tribe, who was Slavic. By the end of the millennium (1000 CE), the area in question would have been unquestionably Slavic.

Helmold:
"Beyond the calmly flowing Odere, and beyond various Pomeranorum peoples, on the Western side we come to the country of the Winulorum, those namely that are called Tholenzi or Redarii. Their city is named Rethre (http://www.jassa.org/?p=1361), a hive of idolatry. There a great temple was built to honor idols, among whom the greatest is Redegast. His statue has ornaments of gold and his bed is bedecked with scarlet. The town itself has nine gates and is surrounded by a deep lake. A wooden bridge is used to traverse it but which is only made available to those who bring offerings or those seeking answers/oracular advice.

(...)

Rani that is Rugiani are a mighty Slavorum people who alone [among the Slavs] has a king; without talking with these people, no public matter may be undertaken for they are afraid of them [this people] on account of the favor that they [Rani] have with the Gods or rather demons, who they worship. These are the peoples of the Winulorum (*) who are spread out among the different regions and provinces and islands of the sea. All those peoples who are dedicated to worshipping idols, always on the move and moving about, engaged in sea piracy.

(...)

After having traversed many Slavorum countries, they came to these who are called Rani or Rugiani and who live in the middle of the sea. There is the hive of errors, there is the capital of idolatry (...) There all the Slavorum come to seek answers/auguries, there they send offerings every year. Merchants who happen to arrive at these places do not receive permission to trade [to buy/sell] until they first make offerings to this God of some expensive items from their merchandise – only then can they display their wares publicly in the market. Their priest they honor no less than [their] king. From that time, when they first fell away from the faith, this superstition has lasted among the Rani to this day.

(...)

These Rani who are also called by some Runi, they are a cruel nation, living in the middle of the sea, and especially given to idolatry. They lead all the Slavs/are the most important of all the Slavorum, they have a king and a most famous temple. Thus, on account of the special veneration of this temple, in which veneration they take first place, they impose their yoke on others but themselves refuse it, being safe by reason of the inaccessibility of the place [i.e., their location]; and the people that their arms conquer they force to pay tribute to temple. They respect their archpriest more than their king and where he/fates [archpriest after determining the fates?] tells them so they send their armies. After received victory they take the gold and the silver to the treasury of their God, dividing the rest amongst themselves.

(...)

Among Slavorum Gods of different shapes the one that shines brightest is Zuantevith [Svantevit], the God of the country of the Rugianorum, as more accurate/successful in his answers [i.e., as an oracle]. In comparison with Him, the Others they treat as if they be demigods. Therefore, to show special veneration for Him, they sacrifice, usually once a year, one Christian chosen by fate. Besides this, from all of the Slavorum countries they send there [to Ruegen/Rugia] offerings such as they determine [in those countries or maybe at the Svantevit temple?]. With special veneration and diligence do they care to serve the temple itself. For they do not easily permit oaths and do not permit people to enter the circle of the temple even when enemies should come.

(...)

In those days Waldemarus [the first], the Danorum King assembled a great army and many ships, so as to set out to the country of the Rugianorum, which he desired to bring under his power. He was helped by dukes Kazemarus and Buggezlavus, the Pomeranorum dukes as well as by Pribislavus the Obotritorum duke – for the duke [Saxon] ordered the Sclavis to help the Danorum King, wherever he should turn his arm to conquer foreign nations. And the Danorum King was lucky and, with a strong hand, he established his rule over the Rugianorum and so as to save their lives, they gave everything to him that he demanded. He ordered then to bring forth that certain ancient statue of Zuantevith [Svantevit], which was worshipped by all the Sclavorum peoples and ordered to put a leash/rope on his neck and to pull it between the warriors [his warrior] to be witnessed by the eyes of the Sclavorum and after having it chopped to pieces to toss into the fire. He destroyed too the temple with all of its holy relics and looted its rich treasury. He ordered the Slavs too to rid themselves of their errors in which they were born and to worship the true God."

http://www.jassa.org/?p=1460

http://www.anthrogenica.com/showthread.php?8006-Rugia-quot-skansen-quot-s%C5%82owia%C5%84ski-R%C3%BCgen-culturally-germanized-genetically-Slavic

* Adam of Bremen:
"Slavania is a very large province of Germaniae inhabited by the Winulis who at one time were called Wandali. It is said to be ten times larger than our Saxonia, especially if you count [as part of Slavia] Boemiam and the expanses across the Oddoram, the Polanos, because they differ neither in appearance nor in language. Although this region is very rich in arms, men and crops, it is shut on all sides by fast barriers of wooded mountains and rivers. In breadth it extends from south to north, that is, from the Albia River to the Scythicum Sea. And in length it appears to stretch from our diocese of Hammaburgensi, where it begins, toward the east and, spread in boundless expanses, reaches clear to Bulgariam, Ungariam and Graeciam.

(...)

For the one, that is the Oddoram, tending toward the north, passes through the midst of the Winulorum peoples."

http://www.jassa.org/?p=1788

08-25-2016, 02:09 PM
I was reading that there is R1a1a cluster in Norway, Maybe these are descendants of the follows of Harald Hardrada, who spend time in exile with the Kievan Rus, when he went back to Norway to claim the throne of Norway, no doubt he took many warriors with him, also, this could be how come there is R1a1a got to Norway, or/and Slave trading? and ultimately how this Haplogroup got to the UK? any thoughts?

Incidentally I think Harald Hardrada was buried in Trondheim; Mary Church until 12th century, this is the exact area where the cluster is.

Baltimore1937
08-25-2016, 05:10 PM
I have Arlogia Vladimirovna, born in Kiev, Kievan Rus, dates conflicting but c.1011 - 1046. She married Rognvald Brusesson who was in temporary exile in Kiev. Otherwise he was the Jarl of Orkney. She became Countess of Orkney. The Bruce line of Scotland descends from them, via Normandy. But don't ask me to prove all of this, ha ha.

Artmar
08-25-2016, 07:46 PM
I was reading that there is R1a1a cluster in Norway, Maybe these are descendants of the follows of Harald Hardrada, who spend time in exile with the Kievan Rus, when he went back to Norway to claim the throne of Norway, no doubt he took many warriors with him, also, this could be how come there is R1a1a got to Norway, or/and Slave trading? and ultimately how this Haplogroup got to the UK? any thoughts?
R1a was present in Norway since Bronze Age and was introduced by people belonging to Battle-Axe culture of a Corded Ware cultural horizon: https://en.wikipedia.org/wiki/Corded_Ware_culture

However, in the most likely scenario R1a people were some time later pushed North to Norway and Northern Sweden by R1b folk. Therefore, they didn't contribute much to proto-Germanic Jastorf Culture and Migration Period. Otherwise, it should be present in higher numbers in Southern Europe and Continent.

R1a started to spread throughout whole Scandinavia probably around Vendel Period: https://en.wikipedia.org/wiki/Vendel_Period and expanded outside Scandinavia with Vikings during Viking Age: https://en.wikipedia.org/wiki/Viking_Age . So it got to Britain just around 8th-11th century for the most part.

So as you can see, Scandinavian R1a is mostly pretty different from Balto-Slavic kinds and only distantly related to them.

I also tend to think that "Slavic" clades of R1a which sometimes pop-up in Norway and Sweden were mostly introduced by Germans (what's sometimes evident going by surnames like Heitmann or Fleischer) and by Slavic craftsmen or mercenaries, rather than slaves. Slaves had a lot less opportunities to left living descendants, although I can imagine some certain cases.

Baltimore1937
08-25-2016, 09:59 PM
There are two different R1a clusters or clades in Norway/Sweden/Finland. The earliest arrival was my L664. Then came the Germanic Z283 > Z284. Z284 is there, but earlier branched off from Z283. In my little subgroup within L664 is a person from southwestern Finland.

Brent.B
10-02-2016, 05:49 AM
Apparently YP4647 formed 1500 years ago (according to Yfull). Assuming these numbers don't change as we get more samples, then how could L1029 get to England by 500AD? Anglo Saxons?

Brent.B
10-08-2016, 06:34 AM
Apparently YP4647 formed 1500 years ago (according to Yfull). Assuming these numbers don't change as we get more samples, then how could L1029 get to England by 500AD? Anglo Saxons?

Does anyone have any thoughts on how reliable this 1500 year estimate is?

Michał
10-08-2016, 09:50 AM
Does anyone have any thoughts on how reliable this 1500 year estimate is?
I suspect this is an overestimation. I will explain it in detail as soon as I find a free minute.

Brent.B
10-11-2016, 07:17 PM
I suspect this is an overestimation. I will explain it in detail as soon as I find a free minute.

I look forward to reading it!

Im not sure if this is just a coincidence, but I noticed that both my sample (YF01875) and sample YF07077 have higher than average SNP counts. The average "corrected" number of SNPs downstream from L1029 appears to be 18.4 (according to Yfull). My sample has 30.13, and YF07077 has 22.31.

I remember reading somewhere that SNPs don't actually occur at a constant rate (genetic mutations are random), but rather tend to approach a uniform rate when applied over a very long period of time (I believe Yfull uses 144.41 years per SNP).


In other words, 2 SNP mutations could occur in one generation, and none in the next. But in the long run, they equal out to give us a steady rate of years per mutation.

If that is the case (Im not sure if it is, I am no expert), then when there is a bottleneck there is a possibility that someone in the bottleneck had a higher/lower mutation rate than normal, leading to an inflated/deflated number of SNPs in the bottleneck .

And that brings me back to this age estimation. We both show a higher than average number of SNPs downstream of L1029. I wonder, then, if there was a higher-than-average SNP count in the time between L1029->YP4647 (the 12 SNPs YF01875 and YF07077 share) due to what I said above.

If we take the 4 SNP difference away from the bottleneck that YF01875 and YF07077 share (we are both at least 4 SNPs above average), then the new corrected number of SNPs for the two would be 26.13 and 18.31 respectively. That would also mean that we share 8 "real" SNPs (not 12).

The age of YP4647, therefore, would be (12+6)/2 (the average SNPs downstream of YP4647) divided by [(12+6)/2]+8 (the total number of "corrected" SNPs under L1029->YP4647).

The result would give us .529, meaning we shared a common ancestor 53% of L1029's age ago. If L1029 is 2100 years old, that means we had a common ancestor 1100 years ago. If we assume an older age of 2400 years for L1029 (Im not sure if this is still your belief Michał?) then we shared a common ancestor 1270 years ago.


of course, we only have 2 samples so these numbers could obviously change easily. If my theory above is correct, then I believe we will continue to see elevated SNP counts for the samples of YP4647. This would also explain my samples very abnormally high SNP count (at least partially).

Does this make sense? Or am I doing/assuming something incorrectly?

lgmayka
10-11-2016, 07:57 PM
Does this make sense? Or am I doing/assuming something incorrectly?
Here's another way to compute it.

Go to the R-L1029 node, and click on its Info. On the TMRCA tab, scroll down to the R-YP4647 branch ID. Its "Age by this line only" is 3229 years, whereas the average age for all of R-L1029 (at the top of the window) is 2089 years.
Now go to the R-YP4647 node, and click on its Info. At the top of the window, the age is listed as 1530 years.

One could reasonably say that: A better estimate of the TMRCA of R-YP4647 is 2089 * 1530/3229 = 990 years. However, Michał would add 10-20% to this, yielding 1100-1200 years.

Brent.B
10-11-2016, 08:20 PM
Here's another way to compute it.

Go to the R-L1029 node, and click on its Info. On the TMRCA tab, scroll down to the R-YP4647 branch ID. Its "Age by this line only" is 3229 years, whereas the average age for all of R-L1029 (at the top of the window) is 2089 years.
Now go to the R-YP4647 node, and click on its Info. At the top of the window, the age is listed as 1530 years.

One could reasonably say that: A better estimate of the TMRCA of R-YP4647 is 2089 * 1530/3229 = 990 years. However, Michał would add 10-20% to this, yielding 1100-1200 years.

Yes, that makes sense.

But would 3229 be the right number to use for age of L1029 (based off of YP4647)? Assuming there is an inflated number of SNPs between YP4647 and L1029 (4 SNPs from my previous post), then the SNP count should drop to 18 from 22 (21.95) for age calculation purposes. Therefore, the 3229 turns into 2600 (18*144.41).

That would make the equation you posted 2089 * 1530/2600 = 1230 years ago.

I just think its strange both my sample and YF07077 have relatively high number of SNPs compared to the average. We know that L1029 isn't actually 3229 years old, so that means that somewhere there are more SNPs than there should be. The question in my mind is did they occur before or after YP4647. Because both samples show higher than average SNPs, I am lead to believe some of the 12 SNPs we share are some of the inflated ones.

Michał
10-13-2016, 01:36 PM
I look forward to reading it!
I am very sorry for responding so late.


Im not sure if this is just a coincidence, but I noticed that both my sample (YF01875) and sample YF07077 have higher than average SNP counts. The average "corrected" number of SNPs downstream from L1029 appears to be 18.4 (according to Yfull). My sample has 30.13, and YF07077 has 22.31.
These are SNPs downstream from CTS11962 (not from L1029).



And that brings me back to this age estimation. We both show a higher than average number of SNPs downstream of L1029. I wonder, then, if there was a higher-than-average SNP count in the time between L1029->YP4647 (the 12 SNPs YF01875 and YF07077 share) due to what I said above.
I agree that the number of SNPs between the L1029 and YP4647 nodes seems to be higher than one should expect based on the data we have for all L1029 members. However, this seems to be only one part of the problem, as it seems to me that what counts the most in this particular case is that your number of private SNPs is strongly skewed (or inflated), as well. I will discuss this below.



If we take the 4 SNP difference away from the bottleneck that YF01875 and YF07077 share (we are both at least 4 SNPs above average), then the new corrected number of SNPs for the two would be 26.13 and 18.31 respectively. That would also mean that we share 8 "real" SNPs (not 12).
I don’t think it is appropriate to call those hypothetical 8 SNPs “real”, as it is rather those 12 SNPs shared by YF01875 and YF07077 that are real in this case. ;)



The age of YP4647, therefore, would be (12+6)/2 (the average SNPs downstream of YP4647) divided by [(12+6)/2]+8 (the total number of "corrected" SNPs under L1029->YP4647).
This does not take into account what I wrote above, ie. that your number of private SNPs (under YP4647) is also strongly inflated.

Let me first notice that although you show 12 private SNPs (all located in the comBED region), the YFull estimates are based on an assumed number of 13 SNPs under YP4647. This is because one “unnamed” known SNP is added to the non-shared novel variants in each such case. This also suggests that YF07077 shows 5 (not 6) private SNPs in the comBED region. I won’t discuss whether this procedure/method used by YFull is justified or not, but this is just to make sure that we use the same procedure when performing such calculations. (BTW, in case you have a contact with YF07077, could you please ask him to join the R1a group at YFull, so our admin team would get access to his YFull data?)

When using your “corrected” number of 8 (instead of 12) SNPs from the YP4647 block, then you show 8+13=21 mutations under L1029 (when using the YFull protocol), which is still much more than the average for L1029 (14.05) not to mention L1029(xYP4647) (13.26), while the corresponding number for YF07077 (8+6=14) fits that average very well. This supports my earlier suggestion that your number of private SNPs is strongly skewed, and you show about 7 “extra” SNPs under YP4647 that make your lineage (and the entire subclade YP4647) look much older than it really is.



The result would give us .529, meaning we shared a common ancestor 53% of L1029's age ago. If L1029 is 2100 years old, that means we had a common ancestor 1100 years ago. If we assume an older age of 2400 years for L1029 (Im not sure if this is still your belief Michał?) then we shared a common ancestor 1270 years ago.
When additionally correcting for your inflated number of private SNPs, the above ratio would drop down to 0.429 (6/14), which would then suggest that clade YP4647 is about 900 years old (when assuming L1029 is 2100 years old) or about 1030 years old (when assuming L1029 is 2400 years old). This is more or less consistent with the estimates calculated by lgmayka, and I think his approach makes perfect sense, as well.



of course, we only have 2 samples so these numbers could obviously change easily. If my theory above is correct, then I believe we will continue to see elevated SNP counts for the samples of YP4647. This would also explain my samples very abnormally high SNP count (at least partially).
Investigating more members of clade YP4647 is indeed the only way to get more accurate estimates. In fact, we already have one more YP4647 member tested with Big Y, so it would be extremely useful to have his BAM file data analyzed at YFull. Importantly, the data extracted from the FTDNA Big Y reports by our admin team seem to support my suspicion that it is your own result that contributes to the inflated age of YP4647 most strongly. When counting all reliable SNPs under YP4647, your number is 13, while kits 413202 and 71694 show only 7 and 5 SNPs, respectively.

Michał
10-13-2016, 01:37 PM
One could reasonably say that: A better estimate of the TMRCA of R-YP4647 is 2089 * 1530/3229 = 990 years. However, Michał would add 10-20% to this, yielding 1100-1200 years.
The fact that YFull adds one extra SNP to the number of private SNPs when calculating the TMRCA age estimates makes me suspect that their TMRCA estimates for relatively young subclades (especially those <1000 years old) are less likely to suffer from the above-mentioned effect seen for older clades.

Brent.B
10-13-2016, 11:33 PM
~

Thank you for the detailed response!

I agree that my sample likely has an overinflated SNP count downstream of YP4647.

Łukasz over at the M458 Facebook group made a graph that visualizes the branch and how it separated.
12131

I do have a question though.

If we assume that YF07077's branch is correct (6 SNPs downstream of YP4647, and 8 upstream as you said), then my sample would have double the SNPs compared to them.

How likely is that? how likely is it that one sample would have 6 extra mutations in a 1000 year timespan?

Could it be possible that the YF07077 branch is slightly under inflated (possibly in their bottleneck?) and my sample is slightly overinflated leading to a larger difference between us?

I hope that question makes sense..




EDIT:

Im not sure if this is the best way to go about answering my question, but I looked through Yfull and found a few haplogroups with a TMRCA very close to 1000 years (YP1368,YP1171,Y11546,S6821,YP412,YP327,YP4454,YP4 078,YP380,YP1361). I took every sample under them and plugged the SNP count for each into Excel.

I found a mean of 6.25 SNPs, and a standard deviation of 1.99. This is with 42 "samples".

That would mean that there is a 95% chance that any sample under a haplogroup that is 1000 years old has between 10.24 to 2.25 SNPs. (6.25 +/- 2*STD)

At 3 standard deviations (99.7% confidence), the range for a 1000 year old haplogroup is 12.24 and .25 SNPs.

So, if YP4647 is actually 1000 years old, then my sample is very unlikely. the highest SNP count found in a 1000 year old haplogroup within 3 standard deviations would be 12.24 SNPs. That makes my sample (of 13 SNPs according to the "NUMBER OF SNPS" bar where I got the information for my calculations) less than .3% likely to happen.

Alternatively, if my sample is overinflated and the other sample is under inflated (meaning the "corrected" number of SNPs downstream of YP4647 falls somewhere in-between 7 and 11), then both samples could be made to fit within 2 standard deviations of the mean. This would also mean that we would need to "correct" the number of SNPs between L1029 and YP4647 by more than four (maybe 5 or 6? no clue). That, of course, would change the ratio closer to 50% if not more

Michał
10-14-2016, 10:26 AM
If we assume that YF07077's branch is correct (6 SNPs downstream of YP4647, and 8 upstream as you said), then my sample would have double the SNPs compared to them.
How likely is that? how likely is it that one sample would have 6 extra mutations in a 1000 year timespan?
As your calculations show, this is relatively unlikely but certainly possible.


Could it be possible that the YF07077 branch is slightly under inflated (possibly in their bottleneck?) and my sample is slightly overinflated leading to a larger difference between us?

Again, this is certainly possible but I would consider it slightly less likely than the above-mentioned scenario, and this is because 1) YF07077 shows more SNPs under L1029 than any other L1029 member (except you), which makes any such hypothetical deflation event less likely, and 2) there are two YP4647 members who show far fewer SNPs under YP4647 than it was found in your lineage (which makes a single event of strong inflation to be slightly more likely than two "independent" events of strong deflation). On the other hand, the lineages represented by 413202 and 71694 are not really "independent", as they both are a part of a common subclade under YP4647 (ie. YP4647>YP4971), so such a deflation could indeed have happened before these two lineages diverged from each other (thus affecting most YP4971 members). For this reason, it is crucial to identify more YP4647* lineages and test them all with NGS/Big Y.

Michał
10-14-2016, 12:26 PM
How likely is that? how likely is it that one sample would have 6 extra mutations in a 1000 year timespan?

When assuming you are right and YF7077 shows a deflated number of SNPs under YP4647, this would mean that the true age of YP4647 should correspond to an intermediate number of SNPs under YP4647 (let's say 8-9 SNPs). However, since we should also keep in mind that the average number of SNPs under L1029 for all remaining L1029 subclades is 13, this would imply that the distance between the L1029 and YP4647 nodes should correspond to about 4-5 SNPs (instead of 12), which in turn would suggest that the lineage ancestral to all YP4647 members accumulated extra 7-8 mutations (so together three times more than expected) within 735-840 years (thus more mutations/year than in your above example).



That makes my sample (of 13 SNPs according to the "NUMBER OF SNPS" bar where I got the information for my calculations) less than .3% likely to happen.

When knowing that 875 members of our project have ordered Big Y so far, we should expect to find individual cases for which the likelihood is about 0.1%.

Brent.B
10-14-2016, 11:58 PM
When assuming you are right and YF7077 shows a deflated number of SNPs under YP4647, this would mean that the true age of YP4647 should correspond to an intermediate number of SNPs under YP4647 (let's say 8-9 SNPs). However, since we should also keep in mind that the average number of SNPs under L1029 for all remaining L1029 subclades is 13, this would imply that the distance between the L1029 and YP4647 nodes should correspond to about 4-5 SNPs (instead of 12), which in turn would suggest that the lineage ancestral to all YP4647 members accumulated extra 7-8 mutations (so together three times more than expected) within 735-840 years (thus more mutations/year than in your above example).


Yes, I was thinking about that as well.

However, when we assume 7 SNPs upstream and downstream of YP4647 with the "corrected" ratio of .5 then the numbers make sense.

That would imply overinflation of 5 SNPs upstream, and 5 SNPs overinflated downstream in my sample. This would also imply 1 SNP under inflation downstream of YP4971.

Working with the numbers I had from before (1000 year haplogroup has a mean of 6.25 SNPs downstream and a standard deviation of 1.99), then a variation of 5 SNPs above and below YP4647 would fall within 3 standard deviations of the mean of 6.25 SNPs for a 1000 year time period. (the range is 12.24 to .25 SNPs).

If we push for any more overinflation in any direction (upstream or downstream of YP4647) then the outcome moves further away from the mean and becomes very unlikely (falls outside 3 STDevs)

Some time this weekend Ill probably gather a much larger sample of 1000 year old SNP counts on YFull and see if the numbers change. Ill also try to email YF07077 to see if he is willing to join the project.


For this reason, it is crucial to identify more YP4647* lineages and test them all with NGS/Big Y.

I 100% agree.

Michał
10-15-2016, 11:09 AM
However, when we assume 7 SNPs upstream and downstream of YP4647 with the "corrected" ratio of .5 then the numbers make sense.
Agreed.


Ill also try to email YF07077 to see if he is willing to join the project.

He has just responded to my message and expressed his willingness to join the R1a group.

Brent.B
10-16-2016, 10:32 PM
Agreed.


Michał, if you still hold the view that L1029 is slightly older than YFull estimates (2400 years), then this would imply YP4647 is 1200 years old.

Assuming that the founder of the YP4647 subclade lived in England, then that would mean the latest possible entry into for YP4647 England was 800AD.

This does not, however, mean that YP4647's ancestors couldn't have arrived there before him. And How likely is it that of the 3 samples we have collected so far, that we have already discovered the earliest branching of the subclade as it came to England?

Would a TMRCA this early indicate that YP4647's ancestors likely came to England slightly before 800AD?

Michał
10-17-2016, 08:01 AM
Michał, if you still hold the view that L1029 is slightly older than YFull estimates (2400 years), then this would imply YP4647 is 1200 years old.

Assuming that the founder of the YP4647 subclade lived in England, then that would mean the latest possible entry into for YP4647 England was 800AD.

This does not, however, mean that YP4647's ancestors couldn't have arrived there before him. And How likely is it that of the 3 samples we have collected so far, that we have already discovered the earliest branching of the subclade as it came to England?

Would a TMRCA this early indicate that YP4647's ancestors likely came to England slightly before 800AD?
We don't know whether the discussed ratio is exactly 0.5 or maybe a bit lower (or higher) than that, or whether the TMRCA age for L1029 is exactly 2400 years or maybe a bit lower, so I wouldn't put too much trust in these very rough estimates. However, based on what we know about the modern distribution (and ethnic association) of clade L1029, it seems more likely that the TMRCA age for YP4647 is younger than 1200 years. Also, even if YP4647 is indeed 1200 years old (or a bit older), this won't rule out its more recent arrival to England, as those hypothetical first YP4647 newcomers could have been cousins (let's say 1st or 2nd cousins), so their MRCA could have predated that westward migration by a couple of decades/centuries.