95% of DYS492=13 in R1b=U106 subhaplogroup

General discussions regarding DNA and its uses in genealogy research

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Tue Jan 24, 2017 6:59 pm
Dr Iain McDonald on his age analysis for U106.
"1. The raw CSV report is compiled, showing the compiled positive and negative calls in a matrix for each person, for each SNP.
2. Missing implied positives are inserted (e.g. if you're L48+ you must be Z381+) and tweaks are made to identify mislabelled insertions/deletions, correct reference negatives etc.
3. The report is sorted horizontally and vertically, to produce the report.csv file that I link to in each update.
4. The report is scanned for shared SNPs, and these are used to form the tree structure. (This is what you see in the HTML report).
5. SNPs are checked to see if they fall within a stable region of the chromosome. They are discarded if not. This reduces the test to about 6.5 million base pairs.
6. The number of SNPs in each node of the tree structure are counted (e.g. U106 = 1 SNP; Z18 + 17364720 + Z19 + Z370 = 4 SNPs).
7. A simple age estimate is produced, purely for checking purposes. This simple age is calculated as:
Number of SNPs / number of tests / coverage / mutation rate = Age
So for U106:
21,587 SNPs / 866 tests / 6,519,270 base pairs / 8.183x10^-10 SNPs per base pair per year = 4673 years per test = 4673 years old
8. Unfortunately, this doesn't take into account many different aspects of the true situation, since not every test is independent. A more detailed calculation is needed. This calculation uses Poisson distribution functions, which describe how many of a rare event you expect to happen from a given sample. The programme loads in a table of distribution functions, which takes the form:
"What is the probability of having <X> SNPs if you would normally expect <Y> SNPs in that length of time?"
In Excel (other spreadsheet software is available), this is equivalent to =POISSON(X,Y,0), which I have tabulated for different X and Y.
9. In ten-year intervals, I compute this probability for each clade. This gives me a distribution of the probability that clade formed at any given date.
10. I then do the same for the parent clade, taking this probability distribution into account (i.e. the parent clade must be older than the child clade).
11. In this way, I build up the tree to the top SNP (U106, in our case).
12. We can then say that the child clade must be younger than the parent, so I multiply together these distribution functions back down the tree to reach the bottom.
13. I then convolve in the Gaussian probability distribution function for the mutation rate, which accounts for the uncertainty in this.
14. I then convolve in the Gaussian probability distribution for the zero point (1950 AD +/- 16 years; 95% c.i.).
15. The final list of ages are tabulated and printed out.

If you are interested in the details, the code is available in the BigY folder on the website. (I need to update this to the latest version).

YFull does exactly the same as this, with a few exceptions:
* In step 5, because they are working directly from the BAM file, they can obtain a little over 8 million base pairs.
* They use a very, very slightly different mutation rate (they use 8.178x10^-10, which is 0.06% faster).
* They don't take into account the full Poisson distribution functions when calculating the ages of clades: they just do a straight average.
* They don't perform the reverse constraint in step 12.
* They don't report actual dates, only clades, so they miss the uncertainty in step 14.
Plus, of course, we have the more complete dataset, despite the loss of nearly 2 million base pairs. The biggest difference for the older clades is in taking into account the full Poisson distribution function, rather than doing a straight average. The biggest differences for the younger clades is in having more testers, and constraing the age from the parent (step 12).

The result is that I expect my ages to be more accurate, because propagating the full probability distribution functions makes a considerable difference. Take U106, for instance (see https://www.yfull.com/tree/R1b/ ). The age of U106 quoted by YFull appears to be 4900 years, but if you click on "info", it actually gives: "R-U106 (age: 4380 ybp)". At the bottom, it says: "NOTE: Age estimation has been taken from downstream subclade R-S263 [R-Z381], its age estimation is more (4900 > 4380)". This problem arises because they take the straight average of all the U106 sub-clades. They give equal weight to clades with only one or two testers as they do to Z381, despite the fact that the age for Z381 is much better determined.

However, I probably over-estimate all the ages slightly, because I can't apply the constraint in step 12 to U106 itself. This is a major reason why I want to build up as far as M269 if I can. Unfortunately, the amount of RAM required for this is rather more than my little laptop can take. It needs at least 32 GB, and there are other issues when dealing with datasets this size.

In the case of YFull, my ages, and any other age estimate, it is very important to pay attention to the range of dates that are quoted. The true date could lie anywhere in this range. For example, I quote an age of 4925 years for Z156. YFull quotes 4487 years. These might seem like two different dates. But if you have a look at the 95% confidence intervals, I quote between 4356 and 5571 years, and YFull quotes between 4100 and 5100 years. Both ranges overlap the others' estimates, which they should do 90% of the time, so the dates agree."

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Mon Feb 06, 2017 3:14 pm
Living Dna state on their page about U106 that the first U106 reached Britain in the Neolithic, so U106 was not carried into west Europe with the Yamnaya Bronze-Age expansion. U106 was already there which we have been saying for the past number of years. The Living Dna company must have some evidence to support the view that was stated.

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Wed Mar 01, 2017 9:00 pm
Was the origin of U106 in Germany 5,000 ybp???? The answer is simple, start digging and testing. That is the only way to find out.

Another reality is that U106 was in western Europe at the time of the Bell Beaker Cultures, Tumulus Cultures and the Urnfield Cultures and also belonged to the Short Neck Beaker culture of eastern Scotland who migrated from the Low Countries to Scotland in the BA. ;)

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Sat Mar 25, 2017 9:31 am
Congratulations to Dr Iain McDonald and the U106 dna project is lucky to have such a highly qualified man on board.
"Strathclyde University appoints genealogy research fellow
A press announcement from the University of Strathclyde's Genelogical Studies Department (http://www.strath.ac.uk/studywithus/cen ... genealogy/):

Honorary Research Fellow Appointed

Dr Iain McDonald has been appointed as Honorary Research Fellow in the Genealogical Studies Department, University of Strathclyde.

Iain comes originally from an Aberdonian family, and began his interest in genealogy 15 years ago, whilst trying to identify any family connection to the Lords of the Isles. Unfortunately, there was no connection, but the process led to an avid interest in Scottish genealogy, and the early history and movement of the Scottish people.

By day, Iain is an astrophysicist, working at the University of Manchester. By night, he has been using physical, statistical and mathematical techniques to develop tools, for both conventional and genetic genealogy.

His speciality is in estimating the dates of male-line (surname) relationships from commercial genetic tests, and using these to construct models of population movements between mainland Europe and the British Isles over the last 5000 years.

Tahitia McCabe, Course leader said: “we are delighted to have Iain on board and look forward to utilising his expertise and knowledge of the technical aspects of DNA testing as applied in genetic genealogy”.

For further information on our study pathways including an online 8-week Introduction to Genetic Genealogy course, please contact us at scosh@strath.ac.uk"

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Sat Apr 08, 2017 7:02 am
Dr. Iain Mc Donald
"There are a variety of options, centred around a theme, regarding the arrival of our ancestors into western Europe. Based on the amalgam of ages derived for our clades, the ancient DNA and the corresponding archeaological cultures, we can be fairly confident our ancestors arrived in western Europe in the centuries surrounding 3000 BC, having come from what later became the European part of the Soviet Union. Deriving anything further than that becomes probabilistic in nature: we cannot be conclusive about what did or did not happen, but must make informed speculation based on the few surviving details we can see.

We know that our ancestors came from easternmost Europe. We don't know that they came directly from the Yamnaya. The Yamnaya ancient DNA belonged to individuals closely related to us, but who were not our direct ancestors. They split from us as the L23 level, around 4200 BC, and around 1000 years before the first Yamnaya ancient DNA in the Samara region of the middle Volga basin. The previous 1000 years is unaccounted for in ancient DNA studies in general, hence our ancestry is not tied specifically to the Yamnaya, but to one the cultures they descend from. This is overlooked in both the lay and academic communities. We can say with good confidence that our ancestors were not in Europe in 5200 BC, but the ancient DNA evidence becomes limited after 4240 BC.

The description I gave of the former Soviet Union is very specific. The Urheimat (ancient homeland) of R-M269 and later R-L23 is theorised to have come from a variety of areas in this region, including the Ural mountains, Caucasus and the Ukraine. The finding of R1b1 remains in Latvia shows that we have a lot to find out about the distribution of our ancestors before circa 3000 BC. The ancient DNA simply isn't sampled sufficiently well across the western former USSR to identify the homeland of M269 conclusively.

The transition to Europe presents different problems. We know that all of R-L11 is descended from one man, but one man alone cannot conquer Europe, so why was he so successful? We know he was not alone. There are now a number of European burials sequenced from the start of the 3rd millenium BC, which evidence the arrival of various R1a1 clades into western Europe at the same time (R1a-S224, R1a-Z281). The presence of other R-M269 basal clades in western Europe, some following the same distributions, suggests there were other, less-successful lines came with them. It was something about the later history of our family that made us more successful.

We can't say explicitly that our ancestors came to Europe with the Corded Ware culture. The Corded Ware culture burials we have are mostly R1a. Oblaczkowo, in western Poland, is the only early Corded Ware R1b1 burial. Unfortunately, the sample recovery was poor, so we don't have any further insight into his haplogroup. A contemporary burial in Lánycsók in southern Hungary is R1b (M343) without further differentiation. It isn't clear if either of these are our ancestors, so it isn't clear if our route into Europe was north of the Carpathians or along the Danube. We can only say that R1b appears strongly present in Europe before 2578 BC, hence that our ancestors are likely to have arrived before this date. We lack a full picture of the genetic upheaval at this time, as we do not have ancient DNA from many cultures which immediately preceded (or were contemporaneous with) the Corded Ware culture, including the Globular Amphorae and Baden cultures. We cannot rule out that our ancestors came to Europe with these cultures (although the Baden culture is nominally thought to be indigenous). The few later burials we have in near-eastern Europe come from the Vatya culture of Hungary, and are haplogroup I2a2a.

The majority of the ancient R-M269 results are from the Bell Beaker culture. The migration of this culture and (possibly) its people is complex and subject to interpretation. It appears to have been carried from Iberia to central Europe, dispersed along the major waterways, then undergone a backflow back to the regions of its origin. The converse is also true: all Bell Beaker culture ancient DNA so far appears to be R-M269. In some cases, this can be specifically haplotyped as R-P312, the implication being that adoption of the Bell Beaker culture by the R-P312 ancestor allowed the spread of R-P312 across western Europe. The earliest dated burial (between 2600 and 2500 BC) is at Kromsdorf, in central Germany, and is M269+ but U106-. The (unproven) implication is that it is P312+. The earliest confirmed P312+ burial is in the nearby Quedlinburg, dated between 2431 and 2150 BC. However, various other P312 Bell Beaker culture burials exist along the upper Danube valley, and would likely be dated between 2500 and 2300 BC. Some of these show P312+ and even U152+.

With the sole exception of RISE98 (2275 - 2032 BC, southern Sweden), there are no truly ancient U106+ DNA. This suggests that U106 was slow to propagate, possibly explaining its small population compared to P312. However, this conclusion is at odds with the large number of direct sub-clades of U106 (current 12), which point to it being instrumental in the population expansion that P312 was involved in. The lack of early U106 could represent many things, socialogical, cultural or statistical. However, the location of RISE98 and the lack of U106 in existing Bell Beaker burials likely indicates that the modern distribution of R-U106 is indicative of its origin in north-western Europe.

The relative timing of the arrival of our ancestors into Europe compared to the rate of SNP formation can still be debated. It is still not clear cut whether the large block of SNPs which include L11, L151, P310 and P311 can be split. However, it is clear that the last common ancestor of this clade represents the beginning of a successful, major population expansion. A logical conclusion is that this population expansion co-incides with the arrival of our ancestors in Europe. We cannot state this with absolute certainty, however studies of the invading cultures (Globular Amphorae and Corded Ware) show little discernable difference in the carbon dates derived from this period, indicating a very rapid expansion (within one or a few generations) of these cultures across Europe.

The last common ancestor of L11 could therefore have been born anywhere in the migration from eastern to north-western Europe. If he was born in eastern Europe, he did not leave any surviving family behind. We know he did leave a lot of family, so there cannot be a long consanguinity of our U106 and P312 ancestors in eastern Europe. So either we are looking at the wholesale movement of the entirety of a large family (which becomes increasingly unlikely as generations pass and the L11 population rapidly grows); or L11 formed somewhere on the path to western Europe, and both U106 and P312 are western European.

With these limits, if L11 arrived in Europe with the Corded Ware culture, this basically dictates that the common L11 ancestor lived some time in the period 2900 - 2700 BC. If L11 arrived with the Globular Amphorae culture, the period moves to somewhere around 3100 - 2800 BC. Of course, these are only two out of a number of possibilities."

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Sat Apr 08, 2017 7:10 am
dartraighe wrote:Dr. Iain Mc Donald
"There are a variety of options, centred around a theme, regarding the arrival of our ancestors into western Europe. Based on the amalgam of ages derived for our clades, the ancient DNA and the corresponding archeaological cultures, we can be fairly confident our ancestors arrived in western Europe in the centuries surrounding 3000 BC, having come from what later became the European part of the Soviet Union. Deriving anything further than that becomes probabilistic in nature: we cannot be conclusive about what did or did not happen, but must make informed speculation based on the few surviving details we can see.

We know that our ancestors came from easternmost Europe. We don't know that they came directly from the Yamnaya. The Yamnaya ancient DNA belonged to individuals closely related to us, but who were not our direct ancestors. They split from us as the L23 level, around 4200 BC, and around 1000 years before the first Yamnaya ancient DNA in the Samara region of the middle Volga basin. The previous 1000 years is unaccounted for in ancient DNA studies in general, hence our ancestry is not tied specifically to the Yamnaya, but to one the cultures they descend from. This is overlooked in both the lay and academic communities. We can say with good confidence that our ancestors were not in Europe in 5200 BC, but the ancient DNA evidence becomes limited after 4240 BC.

The description I gave of the former Soviet Union is very specific. The Urheimat (ancient homeland) of R-M269 and later R-L23 is theorised to have come from a variety of areas in this region, including the Ural mountains, Caucasus and the Ukraine. The finding of R1b1 remains in Latvia shows that we have a lot to find out about the distribution of our ancestors before circa 3000 BC. The ancient DNA simply isn't sampled sufficiently well across the western former USSR to identify the homeland of M269 conclusively.

The transition to Europe presents different problems. We know that all of R-L11 is descended from one man, but one man alone cannot conquer Europe, so why was he so successful? We know he was not alone. There are now a number of European burials sequenced from the start of the 3rd millenium BC, which evidence the arrival of various R1a1 clades into western Europe at the same time (R1a-S224, R1a-Z281). The presence of other R-M269 basal clades in western Europe, some following the same distributions, suggests there were other, less-successful lines came with them. It was something about the later history of our family that made us more successful.

We can't say explicitly that our ancestors came to Europe with the Corded Ware culture. The Corded Ware culture burials we have are mostly R1a. Oblaczkowo, in western Poland, is the only early Corded Ware R1b1 burial. Unfortunately, the sample recovery was poor, so we don't have any further insight into his haplogroup. A contemporary burial in Lánycsók in southern Hungary is R1b (M343) without further differentiation. It isn't clear if either of these are our ancestors, so it isn't clear if our route into Europe was north of the Carpathians or along the Danube. We can only say that R1b appears strongly present in Europe before 2578 BC, hence that our ancestors are likely to have arrived before this date. We lack a full picture of the genetic upheaval at this time, as we do not have ancient DNA from many cultures which immediately preceded (or were contemporaneous with) the Corded Ware culture, including the Globular Amphorae and Baden cultures. We cannot rule out that our ancestors came to Europe with these cultures (although the Baden culture is nominally thought to be indigenous). The few later burials we have in near-eastern Europe come from the Vatya culture of Hungary, and are haplogroup I2a2a.

The majority of the ancient R-M269 results are from the Bell Beaker culture. The migration of this culture and (possibly) its people is complex and subject to interpretation. It appears to have been carried from Iberia to central Europe, dispersed along the major waterways, then undergone a backflow back to the regions of its origin. The converse is also true: all Bell Beaker culture ancient DNA so far appears to be R-M269. In some cases, this can be specifically haplotyped as R-P312, the implication being that adoption of the Bell Beaker culture by the R-P312 ancestor allowed the spread of R-P312 across western Europe. The earliest dated burial (between 2600 and 2500 BC) is at Kromsdorf, in central Germany, and is M269+ but U106-. The (unproven) implication is that it is P312+. The earliest confirmed P312+ burial is in the nearby Quedlinburg, dated between 2431 and 2150 BC. However, various other P312 Bell Beaker culture burials exist along the upper Danube valley, and would likely be dated between 2500 and 2300 BC. Some of these show P312+ and even U152+.

With the sole exception of RISE98 (2275 - 2032 BC, southern Sweden), there are no truly ancient U106+ DNA. This suggests that U106 was slow to propagate, possibly explaining its small population compared to P312. However, this conclusion is at odds with the large number of direct sub-clades of U106 (current 12), which point to it being instrumental in the population expansion that P312 was involved in. The lack of early U106 could represent many things, socialogical, cultural or statistical. However, the location of RISE98 and the lack of U106 in existing Bell Beaker burials likely indicates that the modern distribution of R-U106 is indicative of its origin in north-western Europe.

The relative timing of the arrival of our ancestors into Europe compared to the rate of SNP formation can still be debated. It is still not clear cut whether the large block of SNPs which include L11, L151, P310 and P311 can be split. However, it is clear that the last common ancestor of this clade represents the beginning of a successful, major population expansion. A logical conclusion is that this population expansion co-incides with the arrival of our ancestors in Europe. We cannot state this with absolute certainty, however studies of the invading cultures (Globular Amphorae and Corded Ware) show little discernable difference in the carbon dates derived from this period, indicating a very rapid expansion (within one or a few generations) of these cultures across Europe.

The last common ancestor of L11 could therefore have been born anywhere in the migration from eastern to north-western Europe. If he was born in eastern Europe, he did not leave any surviving family behind. We know he did leave a lot of family, so there cannot be a long consanguinity of our U106 and P312 ancestors in eastern Europe. So either we are looking at the wholesale movement of the entirety of a large family (which becomes increasingly unlikely as generations pass and the L11 population rapidly grows); or L11 formed somewhere on the path to western Europe, and both U106 and P312 are western European.

With these limits, if L11 arrived in Europe with the Corded Ware culture, this basically dictates that the common L11 ancestor lived some time in the period 2900 - 2700 BC. If L11 arrived with the Globular Amphorae culture, the period moves to somewhere around 3100 - 2800 BC. Of course, these are only two out of a number of possibilities."


RISE 98 is of no value to the U106 group because his Y line died out. No modern U106 are connected to him, so it does not matter whether or not he was found in CWC or Battle Axe culture. Some people still insist in using this one U106 RISE 98 example to form their own misguided theory about the origins of all of the U106 group. Get with the real picture, U106 was born, originated, or whatever you want to call it in WESTERN EUROPE.

The ancestor of U106 and P312 was one man, P311 one common ancestor, who did not change the the language nor autosomal dna of western Europe. Get with the program. Western Europe is a huge area and looking for the remains of this one man or the remains of his descendants within the first 500-1,000 years is not a realistic proposition.

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Thu May 04, 2017 3:16 pm
A post by Dr. McDonald.

"An abstract was posted to a conference on Archaeology and Goblets which will take place in Kiel on the week of the 17th. The contribution will report on the Bell Beaker culture ancient DNA analysis. A 20-minute talk is scheduled on the morning of Friday 19th, by Iñigo Olalde and David Reich.

When presenting at conferences, there are often two kinds of abstracts: those where the authors should have something to present by the time of the conference, but where at the time of abstract submission they were vague on the details; and those where the authors have final conclusions that are fairly robustly in place. The detail from the abstract suggests the Bell Beaker results fall into the latter category: while the paper is still to appear in print, they seem happy to provide some results at the conference.

It's tempting to read a lot into this abstract, perhaps too much. The paper will likely be focussing on the autosomal DNA above the Y-DNA and mt-DNA, however it is clear that the majority (81/109) of samples are from the British Isles. Assuming they have tried to recover Y-DNA, which isn't a given, I would expect that they have found that the haplogroup R carried by the Corded Ware culture arriving in the British Isles around 2400 BC. Whether that is R-U106, or even R1b, remains to be seen. I would expect a lot of R-P312 in there based on existing British and European archaeological remains and modern distributions.

We can expect roughly half of the samples to be male. There are no clues as to the level of recovery of the ancient DNA signature. We may be unlucky and only get down as far as haplogroup R in many cases; or we may get lucky and have enough coverage to look at contemporary SNPs. Assuming good coverage, we should be able to tell the frequency of major haplogroups in ancient Britain, including U106, to within an uncertainty of about 10%. The modern percentage is about 20%, so we should hopefully start to get an idea of the fraction of U106 that arrived in Britain during the Neolithic, and the fraction that has arrived during the Metal Age.

However, if the majority of U106 arrived into Europe late in history, we may find few (if any) U106 among these results. What we will get is very scant information on the presence of U106 clades. We will not, for example, have enough data to say whether any particular clade was not present in Britain before 1300 BC. Nor should you expect that they have dug up your direct ancestors: with the contemporary RISE98, we still don't know of any living person belonging to that particular U106 sub-clade.

We may also get some interesting results from continental Europe. By implication there are 28 European results, although this may include results that are already published.

With good coverage, we should also be able to better constrain the mutation rates of ancient DNA, therefore also the ages that I am providing for each clade: a significant amount of the uncertainty here comes from the unknown mutation rate.

Overall, this small amount of information tells us that the project seems to have been successful, and will provide us with some very useful results at some point over the next few months. As a community, we will need to be ready to understand these results and include them in our interpretation of the history of U106. However, you shouldn't expect anything directly related to your particular branch of U106, and the full effects of these data will not be propagated through our community until some months after they are published, but we can expect they'll ultimately have a major impact on what we know about the history of U106 in both Britain and Europe."

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Thu May 04, 2017 7:14 pm
dartraighe wrote:A post by Dr. McDonald.

"An abstract was posted to a conference on Archaeology and Goblets which will take place in Kiel on the week of the 17th. The contribution will report on the Bell Beaker culture ancient DNA analysis. A 20-minute talk is scheduled on the morning of Friday 19th, by Iñigo Olalde and David Reich.

When presenting at conferences, there are often two kinds of abstracts: those where the authors should have something to present by the time of the conference, but where at the time of abstract submission they were vague on the details; and those where the authors have final conclusions that are fairly robustly in place. The detail from the abstract suggests the Bell Beaker results fall into the latter category: while the paper is still to appear in print, they seem happy to provide some results at the conference.

It's tempting to read a lot into this abstract, perhaps too much. The paper will likely be focussing on the autosomal DNA above the Y-DNA and mt-DNA, however it is clear that the majority (81/109) of samples are from the British Isles. Assuming they have tried to recover Y-DNA, which isn't a given, I would expect that they have found that the haplogroup R carried by the Corded Ware culture arriving in the British Isles around 2400 BC. Whether that is R-U106, or even R1b, remains to be seen. I would expect a lot of R-P312 in there based on existing British and European archaeological remains and modern distributions.

We can expect roughly half of the samples to be male. There are no clues as to the level of recovery of the ancient DNA signature. We may be unlucky and only get down as far as haplogroup R in many cases; or we may get lucky and have enough coverage to look at contemporary SNPs. Assuming good coverage, we should be able to tell the frequency of major haplogroups in ancient Britain, including U106, to within an uncertainty of about 10%. The modern percentage is about 20%, so we should hopefully start to get an idea of the fraction of U106 that arrived in Britain during the Neolithic, and the fraction that has arrived during the Metal Age.

However, if the majority of U106 arrived into Europe late in history, we may find few (if any) U106 among these results. What we will get is very scant information on the presence of U106 clades. We will not, for example, have enough data to say whether any particular clade was not present in Britain before 1300 BC. Nor should you expect that they have dug up your direct ancestors: with the contemporary RISE98, we still don't know of any living person belonging to that particular U106 sub-clade.

We may also get some interesting results from continental Europe. By implication there are 28 European results, although this may include results that are already published.

With good coverage, we should also be able to better constrain the mutation rates of ancient DNA, therefore also the ages that I am providing for each clade: a significant amount of the uncertainty here comes from the unknown mutation rate.

Overall, this small amount of information tells us that the project seems to have been successful, and will provide us with some very useful results at some point over the next few months. As a community, we will need to be ready to understand these results and include them in our interpretation of the history of U106. However, you shouldn't expect anything directly related to your particular branch of U106, and the full effects of these data will not be propagated through our community until some months after they are published, but we can expect they'll ultimately have a major impact on what we know about the history of U106 in both Britain and Europe."


"We present new genome-wide DNA data from 196 Neolithic and Bronze Age Europeans – the largest report of genome-wide data in a single study to date – and merge it with published data to form a dataset with 109 BBC individuals that provides a genomic characterization of the BBC across its geographic and temporal range."

They have 87 Neolithic dna samples. That will be interesting.

Nick Patterson
" I think you guys overidentify clonal DNA with culture. These Bronze age cultures
are patrilocal and (for instance) we only have samples from Yamnaya at a few sites.
Hard to know the extent of Yamnaya Y-DNA variation."


Patrilocal =relating to a pattern of marriage in which the couple settles in the husband's home or community. That sounds like indigenous Y haplogroups to me. Are we about to see a few surprises ?!!!!!!!!!!!!!!! I hope so.

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Thu May 11, 2017 11:52 am
Dr McDonald.
"The large number of U106 clades (~13 compared to P312's ~40) indicates that U106 underwent the same population expansion at the same time as P312. However, while we now have about 40 pre-historic* P312 results in the literature, we only have two pre-historic U106 results. That means that, for whatever reason, the current ancient DNA is selecting against U106 in favour of P312. As a group, we need to look into the cultures where ancient DNA is and isn't, and look at the archaeological practices surrounding burials in them. If U106 formed at a time when people started cremating their dead in a particular culture, then we're a bit stuffed for finding them!

(* We have a few historic burials too, but by the Roman era, the haplogroups are too far spread to be useful indicators at the U106/P312 level. The mid-1st-millenium BC Ergolding result could easily be U106>Z326 too.)

The three results we do have, between about 1300 BC and 2200 BC, indicate that the continental distribution of U106 is likely to have been roughly the same as the modern distribution: i.e., strong in the western regions of the Corded Ware culture: the eastern North Sea coast, the western Baltic coasts, and inland into Germany. It seems that U106 may have been stuck in one or more of these regions for longer than we anticipated.

The only bracket we really have of when significant migration of U106 to Britain took place was between about 1500 BC and 300 AD. And even then, it seems like later migrations (e.g. Danes, Angles, Saxons, Jutes, Normans) must have brought in a lot more.

In a wider context, the lack of ancient U106 in the Rhine or Danube valleys, and the lack of haplogroup R and Yamnaya at-DNA in the Globular Amphorae culture suggests the route taken by our common P311 forebears was north of the Carpathians, probably with the Corded Ware culture. I keep changing my mind about how secure this result is, as new data comes in, but I think the latest data add a lot of evidence against the Danubian route. The samborek (Poland) results from before 2672 BC may confirm this. One possibility is that P311 was a coastal culture, which passed around the Baltic coast, and didn't really settle down on land until it reached the Danish Isles. I wouldn't say there's any real evidence for this, necessarily, just that it's an option.

Presumably the raw data will become available on full publication on the articles, at which point it will be interesting to see where any P311xP312xU106 results lie, and whether any further sub-clades can be determined. The distribution of S1194 seems to largely reflect the distribution of U106, which is consistent with them both representing the point from which P311 spread. The rapid uptake of P312 into the Bell Beaker culture, and its rapid spread, means that its modern distribution is more representative of its own founder effect, rather than the P311 founder effect.

If, at this point, I had to guess the contribution to Great British U106 by the Angles, Saxons, Jutes, Danes and Flemish between 410 AD and 1066 AD, I'd say it's likely to be of order 50%, perhaps a little more. We see some lines which look older, and some lines we know came in with the Normans. But there are some lines that look early Celtic, then Germanic, then Anglo-Saxon, then Gaelic. A lot of cultural migration has always gone on in short periods of time, so I'm not inclined to stick labels for particular clades against particular cultures, unless the two are roughly contemporary."

Posts: 2223
Joined: Fri Mar 16, 2012 5:43 pm

MtDNA:
U5b2b
PostPosted: Thu May 11, 2017 12:05 pm
An excellent post by Dr. Iain.

"Determining the real origin of clades is extremely difficult, and impossible to do conclusively this far back in history. Inferences must be made based on the time that clade originates, its point of origin based on modern distributions (taking into account founder effects and testing biases as best you can), and known archaeological cultures in the area at the time. At best, this is guesswork. But many of the guesses so far have turned out to be right based on archaeological DNA... not all of them, but many!

The clades I was thinking of are all within U106>Z156>Z304>DF98, which is the one I'm most familiar with. As a whole, the clade looks to be Rhineish is origin, or at least to the southern end of the U106 distribution, and roughly contemporary with the Bell Beaker culture. Look close enough, and you find some interesting clades within this branch.

R-S22069 in particular is very closely connected to the region around Frankfurt. This was a major region where both Celts and their precursors settled, in periods closely connected to the origins of several R-S22069 clades (circa 1600 BC and 600 BC). Within this, we know that R-S8350, the royal Wettin branch, has its origins 1000 years ago in Germanic cultures, and that a branches of this family appear in England from Norman or slightly pre-Norman times, with a later medieval branch turning up in southern Scotland.

Simultaneously, we have R-S1911, which is more widely spread around the extremities of Celtic society, and is particularly strong in the parts of Britain normally shunned by R-U106. The sub-clade R-S1894>S4004>FGC14814 has been in the UK since at least the 3rd century AD, as attested by the Roman burials in York (the clade itself is probably about 7th century BC). Its origin in the UK may well be Celtic or pre-Celtic (but post Bell Beaker). One particular branch has ended up in Scandinavia.

The other major R-S1911 sub-clade, R-S10621, is also found in Celtic extremities of Europe, from Ireland and Scotland, to western Ukraine, to Spain. Its grandson clade, PH589>S17136, however, has an origin in the first few centuries AD, and is found in historically Germanic regions of NW Germany and the Netherlands. One family (Decker) sits right in the middle of where the Saxons lived. English names of this particular sub-clade suggest a southernly distribution, following the Anglo-Saxon-etc. migrations.

The third, minor branch of R-S1911, is probably also found in the western Ukraine, but is better attested in Ireland, and in the Norman family of de Dutton, showing a presence in medieval France.

So what are we to make of all this? The basic point is that history is a mess. You can't take a 5000-odd-year-old clade like U106 and call it "Germanic". You can't take a 4000-odd-year-old clade like DF98 and say it is "Celtic". You can't take a 3000-year-old clade like S22069 or S10621, call it pre-Celtic, and expect all its descendants to be Celts and not Germans. You can't even take a 1000-year-old clade like S8350 is Germanic in origin, ends up being mostly English, and then expect only 300 years later to find its descendants on the same side of the Scottish Wars of Independence.

In 52 BC, Vercingetorix lost the Battle of Alesia to Julius Caesar and, almost overnight, two million (Celtic) Gauls became Romans. They were the same people. They had the same ancestors as before. They (at least initially) followed many of the same customs. But we pigeonhole them differently in history because of who their rulers were and how they were integrated into the world. Imagine this story played out across the millennia, and you'll see how your ancestral line has changed cultures more times than you've changed underwear. Cultural assignations for haplogroups only work when they're done contemporaneously: it can only represent the culture in which that particular clade formed.

Similarly, we can't say that make claims like R-U106 came to Britain with the Anglo-Saxons, or whoever else. Each lineage is unique. Migrations of U106 to Britain probably started shortly after it was formed, and continue to the present day. Within that, there will have been many "unsuccessful" migrations that left no descendants in Britain today: we only see the "successful" ones. Some of these, like the de Dutton family and the British branch of R-S8350, were relatively recent. Some, like R-FGC14814, have been here for much longer. It's important to look at the individual cases."
Top
PreviousNext

Return to General DNA Discussion

Who is online

Users browsing this forum: No registered users and 2 guests