PDA

View Full Version : Great Britain and Ireland Sub-Regions



TopLobster
09-13-2018, 12:58 PM
How is Living DNA able to split Central, East, and South England into multiple sub-regions when the POBI project showed that these areas remained in one cluster even when the number of clusters was increased to 24? See image for what I am referring to.

25884

JonikW
09-14-2018, 10:59 PM
How is Living DNA able to split Central, East, and South England into multiple sub-regions when the POBI project showed that these areas remained in one cluster even when the number of clusters was increased to 24? See image for what I am referring to.

25884

I remember someone posting about this a good while back with a lot more info than I can remember now. But given that no one else has replied, I think Living DNA just claim they've also added their own samples and are overall confident of their ability to differentiate more than POBI dared to do. My own English results certainly look reasonably good (I was in the first batch of testers). At least, they didn't throw in anything unexpected or unlikely such as East Anglian (I'm not from Kent although the southeast has become my home)....

Sikeliot
09-15-2018, 04:27 AM
I am surprised the red "SE England" cluster is homogenous across that entire geographic space. I would expect there to be some regions breaking off from it due to higher Germanic. There is no way to me that somewhere like Hampshire or Berkshire has as much Germanic admixture as does East Anglia or Kent.

sktibo
09-16-2018, 03:23 PM
I am surprised the red "SE England" cluster is homogenous across that entire geographic space. I would expect there to be some regions breaking off from it due to higher Germanic. There is no way to me that somewhere like Hampshire or Berkshire has as much Germanic admixture as does East Anglia or Kent.

This is off the top of my head but it was said that it takes an incredibly close examination to break the large red SE English cluster into separate regions. While some Living DNA test results were able to successfully split the ancestry of some into regions within it for others it was not able to. I personally think the scale is too fine. For all we know someone from a place like Hampshire could have more Germanic ancestry than someone from East Anglia due to later migrations from the continent like France or Belgium. Norfolk's collection of aDNA results reveals a very French-like autosomal picture and for an East Anglian clusters very far from the Germanic types. It's difficult to say with certainty at the time being.

Edit: This youtube video shows and explains how the POBI samples break down into their clusters:
https://www.youtube.com/watch?v=6ANNHMzmxlI
Starts around 16:40
He says the large red cluster never splits even at 53 clusters. I think I remember hearing somewhere that if you go past this level it does eventually split but at that point the differences are incredibly tiny.

The Irish DNA Atlas has one large English cluster (blue on their map instead of read) at K30 but at K48 four English clusters split off from this but they are all very tiny - two have 1 individual and the other two have only 3. So two analyses conclude that it's a pretty darn genetically homogeneous area.

mxcrowe
09-18-2018, 11:49 PM
It may just be coincidental, but the "cautious" settings, where they group certain sections back together, are actually the most accurate reflection of my known genealogy. They put me at:

Northwest England-related ancestry - 42.2%
Northumbria-related ancestry - 22.3%
North Yorkshire - 8.1%
Orkney-related ancestry - 6.7%
Southeast England-related ancestry - 6.4%
and Cornwall and North Wales at around 2% each

*If* they had caught my (high confidence and relatively recent) line from Mecklenburg (Prussia), I would have been truly blown away by how close this matches what I know. However, for whatever reason, the Prussian component appears to hide in the English genes.

So, for what it's worth, and I admit it could just be a happy coincidence, their cautious map is extremely accurate for me, including the strange bridging of areas that, at first glance, one would not necessarily believe are connected.

JonikW
09-19-2018, 08:24 AM
It may just be coincidental, but the "cautious" settings, where they group certain sections back together, are actually the most accurate reflection of my known genealogy. They put me at:

Northwest England-related ancestry - 42.2%
Northumbria-related ancestry - 22.3%
North Yorkshire - 8.1%
Orkney-related ancestry - 6.7%
Southeast England-related ancestry - 6.4%
and Cornwall and North Wales at around 2% each

*If* they had caught my (high confidence and relatively recent) line from Mecklenburg (Prussia), I would have been truly blown away by how close this matches what I know. However, for whatever reason, the Prussian component appears to hide in the English genes.

So, for what it's worth, and I admit it could just be a happy coincidence, their cautious map is extremely accurate for me, including the strange bridging of areas that, at first glance, one would not necessarily believe are connected.

I think my cautious results are best too. Perhaps the broader approach is best for some people in weeding out a bit of noise.

Phoebe Watts
09-19-2018, 08:52 AM
It may just be coincidental, but the "cautious" settings, where they group certain sections back together, are actually the most accurate reflection of my known genealogy. They put me at:

Northwest England-related ancestry - 42.2%
Northumbria-related ancestry - 22.3%
North Yorkshire - 8.1%
Orkney-related ancestry - 6.7%
Southeast England-related ancestry - 6.4%
and Cornwall and North Wales at around 2% each

*If* they had caught my (high confidence and relatively recent) line from Mecklenburg (Prussia), I would have been truly blown away by how close this matches what I know. However, for whatever reason, the Prussian component appears to hide in the English genes.

So, for what it's worth, and I admit it could just be a happy coincidence, their cautious map is extremely accurate for me, including the strange bridging of areas that, at first glance, one would not necessarily believe are connected.

LDNA seems to work best for testers with most of their ancestry in north and west Britain. And cautious tends to bring in some of the Ireland and Scotland that are "missing" in the standard view.

In my case the standard view is best. My cautious view "lost" all the "real" south Wales ancestry that had been identified in the standard view and pushed it into a South Wales Borders-related category that didn't include south Wales. So there are quirks in cautious and I think it is useful look at all three views.

JonikW
09-19-2018, 09:23 AM
LDNA seems to work best for testers with most of their ancestry in north and west Britain. And cautious tends to bring in some of the Ireland and Scotland that are "missing" in the standard view.

In my case the standard view is best. My cautious view "lost" all the "real" south Wales ancestry that had been identified in the standard view and pushed it into a South Wales Borders-related category that didn't include south Wales. So there are quirks in cautious and I think it is useful look at all three views.

I'd like them to do more on Wales. The Borders category focuses on the east where I have Monmouthshire and Breconshire ancestry, while South Wales seems focused on the far west in the POBI dataset at least, which had almost no samples in between, and nothing in the middle of the country.

EDIT: added map

26034

JonikW
09-19-2018, 02:19 PM
How is Living DNA able to split Central, East, and South England into multiple sub-regions when the POBI project showed that these areas remained in one cluster even when the number of clusters was increased to 24? See image for what I am referring to.

25884

Just took another look at Wales. Given that Living DNA hasn't even sorted out Scotland and Ireland, I see no chance of them getting round to improving things though.

Their current categories seem to offer a lot more than POBI itself because the Living DNA definitions cover the whole country and more. Here's a recap:

"The North Wales signature roughly covers the areas of Anglesey, Gwynedd, Conwy, Wrexham and northern Powys.

The unique southern signature is found in the modern counties of Pembrokeshire, Ceredigion, Carmarthenshire and West Glamorgan.

The areas of Shropshire, Herefordshire, Monmouthshire, Worcestershire, Powys and Gwent are collectively called the South Wales border."

Perhaps they've added loads of quality samples as we were discussing at the top of this thread for England? Unlikely I suppose given what we've seen for Scottish and Irish testers.

On their current regions, one thing that is interesting is how the Borders category shows that "Celtic" Welsh ancestry is still high along the English side, as has been pointed out before. It corresponds nicely with the old natural border having once been the Dee and Severn and suggests that this region, in going beyond modern political boundaries, shows real population structure that reflects this.

I would really like to see proper sampling of Mid Wales, taking people who knew their ancestry beyond just four grandparents. Would it form its own cluster(s) or show the current North, South and/or Borders, I wonder. Of course early Powys included the Borders (well into modern England) and much of Mid Wales (map attached). Has anyone tested who can shed some local light?

26040

mxcrowe
09-19-2018, 08:35 PM
I think my cautious results are best too. Perhaps the broader approach is best for some people in weeding out a bit of noise.

Yes, that is my sense of it, too. At what point does all our desire for regional specificity break down into silliness? So, the fact that the "cautious", broadest view fit my known genealogy data the best made sense to me; however, as Phoebe Watts points out: "LDNA seems to work best for testers with most of their ancestry in north and west Britain. And cautious tends to bring in some of the Ireland and Scotland that are "missing" in the standard view.

In my case the standard view is best. My cautious view "lost" all the "real" south Wales ancestry that had been identified in the standard view and pushed it into a South Wales Borders-related category that didn't include south Wales. So there are quirks in cautious and I think it is useful look at all three views." These are good points and part of what worries me about being given, say, three different interpretations so I can pick the one that best fits. It all begins to sink into speculation and stacked interpretations.

Despite the clear patterns that POBI discovered, probably to the surprise of many, at some point you just hit the limitations of meaningfulness when people are free to move around as time goes on. My ancestry shows periods where everyone stayed in the same area for relatively long periods (several generations) punctuated by sudden moves from one region to another, which I'm sure is fairly typical for most of us. I guess what I'm saying is that I think Living DNA's regional classification is at (or even beyond) what is actually meaningful.

JonikW
09-19-2018, 08:49 PM
Yes, that is my sense of it, too. At what point does all our desire for regional specificity break down into silliness? So, the fact that the "cautious", broadest view fit my known genealogy data the best made sense to me; however, as Phoebe Watts points out: "LDNA seems to work best for testers with most of their ancestry in north and west Britain. And cautious tends to bring in some of the Ireland and Scotland that are "missing" in the standard view.

In my case the standard view is best. My cautious view "lost" all the "real" south Wales ancestry that had been identified in the standard view and pushed it into a South Wales Borders-related category that didn't include south Wales. So there are quirks in cautious and I think it is useful look at all three views." These are good points and part of what worries me about being given, say, three different interpretations so I can pick the one that best fits. It all begins to sink into speculation and stacked interpretations.

Despite the clear patterns that POBI discovered, probably to the surprise of many, at some point you just hit the limitations of meaningfulness when people are free to move around as time goes on. My ancestry shows periods where everyone stayed in the same area for relatively long periods (several generations) punctuated by sudden moves from one region to another, which I'm sure is fairly typical for most of us. I guess what I'm saying is that I think Living DNA's regional classification is at (or even beyond) what is actually meaningful.

You make some excellent points mxcrowe. I just don't believe Living DNA is that accurate at a fine level. For me, the most conservative approach is best. Look how broad my map is at the cautious level. It fits everything I know on my paper trail. Even then, I only know the names of less than a third of my ancestors at 64 gggg grandparent level. I recently discovered one from Galway who I had no idea about until then. Who knows what else I might uncover if I keep at it. Moreover, it's remarkable how well cautious mode correlates with the map Lucasz did for me, which utilised my separate 23andme results. I'll post it below too.

ADD: the mere fact they pegged me as predominantly Welsh Borders is astonishing enough given my father's Breconshire and mother's Monmouthshire/Breconshire. Pretty amazing when you consider how close people throughout Europe are to each other...

26047

26048

mxcrowe
09-19-2018, 09:21 PM
You make some excellent points mxcrowe. I just don't believe Living DNA is that accurate at a fine level. For me, the most conservative approach is best. Look how broad my map is at the cautious level. It fits everything I know on my paper trail. Even then, I only know the names of less than a third of my ancestors at 64 gggg grandparent level. I recently discovered one from Galway who I had no idea about until then. Who knows what else I might uncover if I keep at it.

ADD: the mere fact they pegged me as predominantly Welsh Borders is astonishing enough given my father's Breconshire and mother's Monmouthshire/Breconshire. Pretty amazing when you consider how close people throughout Europe are to each other...

26047
Agreed on all points. So much of this comes down to how the recipient of the data interprets it (nevermind all the interpretations already made by the testing company!). Imagine if you were an adoptee in Germany and you received the map you attached - you'd be ecstatic at having narrowed your ancestry down to such a small region and at that level of detail. On the other hand, someone who has a fairly certain genealogy for virtually everyone back to the 1024-level (is there such a person?) will feel like he hasn't learned anything from the same report! To further complicate, maybe the DNA is showing us things that the paper trail has kept secret for generations? I guess it's much like everything else in genealogy: you use every tool at your disposal, uncover every piece of data you can, and hold everything in the "maybe" category until you feel like it's confirmed beyond a shadow of a doubt (should that lucky day ever come...).

FionnSneachta
09-19-2018, 09:24 PM
Cautious mode is definitely more accurate for me for obvious reasons. At least with sub-region cautious mode, I get assigned 61.3% Southwest Scotland-related ancestry which includes Ireland compared to my 19.6% Ireland in standard mode. Regional cautious is even more accurate assigning me 98.2% Great Britain and Ireland with the global being the most accurate at 100% Europe. Of course, British and Irish people don't test with a company like Living DNA to be told that they're from just Europe or just Great Britain and Ireland.

JonikW
09-19-2018, 09:51 PM
Agreed on all points. So much of this comes down to how the recipient of the data interprets it (nevermind all the interpretations already made by the testing company!). Imagine if you were an adoptee in Germany and you received the map you attached - you'd be ecstatic at having narrowed your ancestry down to such a small region and at that level of detail. On the other hand, someone who has a fairly certain genealogy for virtually everyone back to the 1024-level (is there such a person?) will feel like he hasn't learned anything from the same report! To further complicate, maybe the DNA is showing us things that the paper trail has kept secret for generations? I guess it's much like everything else in genealogy: you use every tool at your disposal, uncover every piece of data you can, and hold everything in the "maybe" category until you feel like it's confirmed beyond a shadow of a doubt (should that lucky day ever come...).

Sorry, must confess I added that Lucasz map to my post after you quoted me. It's worth a look. I've just discovered my late mum's first cousin through matching and he's told me some hair-curling tales about illegitimate children and births outside marriage. I don't think anyone can be confident about all their ancestors at gggg level.

mxcrowe
09-19-2018, 10:58 PM
Sorry, must confess I added that Lucasz map to my post after you quoted me. It's worth a look. I've just discovered my late mum's first cousin through matching and he's told me some hair-curling tales about illegitimate children and births outside marriage. I don't think anyone can be confident about all their ancestors at gggg level.

Two of my ancestors, sisters, married two gentlemen. One of the sisters had (knowingly, I believe) married her first cousin, and they had four children, none of whom married or had children themselves. The other couple had no children. While both couples were still quite young, the first-cousin male died and the other sister died, so the remaining sister married her sister's former husband. That couple produced four more children, one of whom is my grandmother, so that's a good thing. Now, the chap that she remarried to (my great-grandfather) lists no father on his birth cert (or on any record I can find) and his surname is his mother's surname, as listed on the birth cert. So I have no known ancestors up that line and don't know the surname to even go looking, and my grandmother's surname is not her (real, traditional?) surname. As complicated as that is, my great-grandmother, the one who remarried her sister's husband, comes from a family line that is not her proper surname either, as her grandfather took his wife's surname upon marriage. So my grandma's surname is not her surname and her mother's surname is not her surname either!

Point is, this is one that I unraveled...just think how many there are like this and worse as you go up the tree!

Sikeliot
09-20-2018, 12:54 AM
On their current regions, one thing that is interesting is how the Borders category shows that "Celtic" Welsh ancestry is still high along the English side, as has been pointed out before. It corresponds nicely with the old natural border having once been the Dee and Severn and suggests that this region, in going beyond modern political boundaries, shows real population structure that reflects this.

It might suggest that the Anglo-Saxon input is relatively limited there, and that the original population in that part of England would have always been close to the people who became the Welsh if not identical to them, so even today, those "English" people in places like Shropshire, Staffordshire, Worcestershire are still close to the Welsh.

JohnHowellsTyrfro
09-20-2018, 09:02 AM
Something similar was recently mentioned in a TV programme on DNA.
The suggestion was that central England has been the focus for migration, mingling and settlement over a very long period. It is the centre of England I suppose.
The more "remote" areas were more able to retain specific genetic identities, dependent on place of course.

Phoebe Watts
09-20-2018, 12:28 PM
I would agree that the LDNA regions are about right. Hopefully the improvements from Irish and Scottish sampling will clarify Welsh results too. What we are missing is the sampling of the borders and parts of rural Wales. And then a historical interpretation. The history of the borders is fascinating.

sktibo
09-20-2018, 03:09 PM
Yes, that is my sense of it, too. At what point does all our desire for regional specificity break down into silliness? So, the fact that the "cautious", broadest view fit my known genealogy data the best made sense to me; however, as Phoebe Watts points out: [I]"LDNA seems to work best for testers with most of their ancestry in north and west Britain. And cautious tends to bring in some of the Ireland and Scotland that are "missing" in the standard view.


Great question, I think with Living DNA's proposed 13 Irish regions and God-knows how many Scottish regions it's certainly getting to be silly. I think they definitely need more of both right now, as Ireland more or less doesn't even have a region with its tiny reference panel and the entire middle of Scotland is unsampled. I also found that my Scottish and Northern Irish ancestry to be more or less missing too - I got an enormous percentage of Northumbria and then 11.5% in Aberdeenshire where I have very little ancestry from, at most 3% and that's probably generous. Somehow zero percent in SW Scotland and NI despite Scotch-Irish ancestry and further ancestry from Ayrshire.. I think you're far from alone with your missing ancestry on this test. The biggest mind boggle of all is absolutely no central English ancestry in my case, with a great-grandmother born in either Derbyshire or Nottinghamshire (can't remember which right now) and all of her ancestry has been traced to those two counties plus Staffordshire until about 1620.. I believe this is my family's longest and best researched line, on top of that the ancestry is relatively recent and there's not a fraction of a percentage.

...

I checked out my cautious mode again and I suppose that if I let go of the regional titles it is a bit more in line with my known genealogy but pretty far off still. In regular mode I get zero NW Scotland which is supposedly where my ancestry from the lowland central belt should land (I find it pretty damn hard to believe that lowlanders from Stirling would actually end up with the samples from Islay, Argyll, and Lewis...) but according to Living DNA's website that's the case... however my Northumbrian cautious group includes NW Scotland at 44.6 which is quite an overshot. If Perthshire and Stirlingshire are included in NW Scotland, that should be about 11% in my case. However, "Orkney-related ancestry" includes SW Scotland and NI, at 13.4% which is pretty close to the expected amount which is again close to 11% - plus there is a bit of Orcadian in me although it is less than 1% - somewhere around 0.75%. So not bad for this category if we count it as SW Scotland/NI. South England Related ancestry (17.3%) does pretty good too: Should be around 12.5%, not unreasonably far off. Unfortunately no matter how I look at it there's still nothing for Central England and nothing for mainland Europe on this mode, which is the source of a lot of my ancestry.. however, if interpreted in this way I think Cautious does better than Standard which assigns me 42% Northumbria... 39% over what it should be.. and that's just one of my percentages. It gets more wrong than right when compared to the paper trail.

avalon
09-20-2018, 06:59 PM
Just to echo some of the thoughts here, my own LivingDNA cautious grouping was pretty good too - it gave me 78.3% NW England related and 15.8% North Wales and 5.8% unassigned. Given that almost all of my known ancestry is from NW England and North Wales, then I can't complain too much, the only real issue is that LivingDNA were unable to distinguish my NE Wales ancestry (Flintshire and Denbighshire) from its North West England cluster. My 15.8% North Wales is almost certainly related to my ancestry from Merionethshire.

By way of comparison, I recently tested with AncestryDNA which has given me 87% England, Wales, NW Europe and 13% Ireland and Scotland. What's interesting is that whilst Ancestry clearly has some labelling issues, there are interesting similarities with my Living DNA cautious view. I was also interested in Phoebe's results (someone who is 100% Welsh), IIRC LivingDNA gave you some unexpected "South Wales Border" in Cautious mode and via Ancestry you were 50% Ireland/Scotland and 50% Eng,Wal, NW Europe?

So, I think there are some interesting parallels between Ancestry and LivingDNA, in my case it looks like a correlation between LivingDNAs North Wales region and Ancestry's "Ireland/Scotland" region, I'd be interested to know if others notice similar patterns.

What we are essentially looking at here is different types of Welsh ancestry -some of which is more distinct and perhaps "Irish" like and other Welsh ancestry that is difficult to separate from English. None of this is of course surprising given the known history of Wales. At the moment though LivingDNA is more specific and accurate than Ancestry, though I think both are broadly saying the same things.

mxcrowe
09-20-2018, 07:44 PM
... if I let go of the regional titles it is a bit more in line with my known genealogy ...

I think this is a key point. Back to interpretations: many of us jump to a specific conclusion when a regional name is used, even though the political borders have been quite fluid over time. While this is less of an issue for those with British Isles heritage, it's enormously complicated in say, northern Italy, which has been under several different political jurisdictions (and gone by several different names) in just the past 200 years. So if we look very generally at geographical areas where populations might have remained somewhat insular, without as much focus on the various names the testing companies come up with when labeling said areas, we'll probably be more accepting of of their interpretations of these being representative of our genetic makeup.

Obviously, places that have been "melting pots", as JohnHowellsTyrfro points out, will be less useful as reference populations.

Phoebe Watts
09-20-2018, 08:21 PM
Just to echo some of the thoughts here, my own LivingDNA cautious grouping was pretty good too - it gave me 78.3% NW England related and 15.8% North Wales and 5.8% unassigned. Given that almost all of my known ancestry is from NW England and North Wales, then I can't complain too much, the only real issue is that LivingDNA were unable to distinguish my NE Wales ancestry (Flintshire and Denbighshire) from its North West England cluster. My 15.8% North Wales is almost certainly related to my ancestry from Merionethshire.

By way of comparison, I recently tested with AncestryDNA which has given me 87% England, Wales, NW Europe and 13% Ireland and Scotland. What's interesting is that whilst Ancestry clearly has some labelling issues, there are interesting similarities with my Living DNA cautious view. I was also interested in Phoebe's results (someone who is 100% Welsh), IIRC LivingDNA gave you some unexpected "South Wales Border" in Cautious mode and via Ancestry you were 50% Ireland/Scotland and 50% Eng,Wal, NW Europe?

So, I think there are some interesting parallels between Ancestry and LivingDNA, in my case it looks like a correlation between LivingDNAs North Wales region and Ancestry's "Ireland/Scotland" region, I'd be interested to know if others notice similar patterns.

What we are essentially looking at here is different types of Welsh ancestry -some of which is more distinct and perhaps "Irish" like and other Welsh ancestry that is difficult to separate from English. None of this is of course surprising given the known history of Wales. At the moment though LivingDNA is more specific and accurate than Ancestry, though I think both are broadly saying the same things.

That's interesting, thanks.

The treatment of my south Wales ancestry in LDNA Cautious is just odd - they identified it correctly in Standard and pushed it east in Cautious. South Wales is from my father's side and it is correct in his Cautious and Standard modes.

LDNA's explanation was that there might be something that looked Germanic Anglo-Saxon in there. There is a pull to the east in some of my K36 results too; Lukasz showed a Dutch influence for me and some Irish testers; a K36 based map placed me in the sea, just north-east of London.

About 50%/50% Ireland&Scotland and England, Wales etc. seems to be common for testers from north-west Wales and my south Wales ancestry doesn't seem to have any effect. I haven't seen many Welsh results after the update though. It would be interesting to see more. Perhaps it is an east/west effect, rather than north/south?

JonikW
09-20-2018, 08:58 PM
That's interesting, thanks.

The treatment of my south Wales ancestry in LDNA Cautious is just odd - they identified it correctly in Standard and pushed it east in Cautious. South Wales is from my father's side and it is correct in his Cautious and Standard modes.

LDNA's explanation was that there might be something that looked Germanic Anglo-Saxon in there. There is a pull to the east in some of my K36 results too; Lukasz showed a Dutch influence for me and some Irish testers; a K36 based map placed me in the sea, just north-east of London.

About 50%/50% Ireland&Scotland and England, Wales etc. seems to be common for testers from north-west Wales and my south Wales ancestry doesn't seem to have any effect. I haven't seen many Welsh results after the update though. It would be interesting to see more. Perhaps it is an east/west effect, rather than north/south?

Very interesting re a possible east/west effect. There's just so much we can't understand at present. I'm east in as much as I show as largely Welsh Borders but that K36-based map you mentioned put me in Snowdonia. I would have thought you were more likely to be there than me given your fully Welsh background. I just don't get any of it...

avalon
09-21-2018, 04:53 PM
That's interesting, thanks.

The treatment of my south Wales ancestry in LDNA Cautious is just odd - they identified it correctly in Standard and pushed it east in Cautious. South Wales is from my father's side and it is correct in his Cautious and Standard modes.

LDNA's explanation was that there might be something that looked Germanic Anglo-Saxon in there. There is a pull to the east in some of my K36 results too; Lukasz showed a Dutch influence for me and some Irish testers; a K36 based map placed me in the sea, just north-east of London.

About 50%/50% Ireland&Scotland and England, Wales etc. seems to be common for testers from north-west Wales and my south Wales ancestry doesn't seem to have any effect. I haven't seen many Welsh results after the update though. It would be interesting to see more. Perhaps it is an east/west effect, rather than north/south?

Can you remind me again Phoebe, what did LivingDNA assign you for North Wales and South Wales in complete mode? Your case is interesting because you have ancestry from north and south.

It's difficult to draw firm conclusions - Welsh results with AncestryDNA and LivingDNA are few and far between but I'm wondering whether what AncestryDNA calls your " 50% Ireland/Scotland" is largely made up of your LivingDNAs North Wales ancestry whereas your South Wales is in Ancestry's England, Wales, NW Europe category. It might not be as clear cut as that though, it could be more of a mixture, your "Ireland/Scotland" might be part South Wales, part North.


Based on my own results I would say that within North Wales there is an east-west split genetically, probably centered around Flintshire and bits of Denbighshire. What I'm not too sure about is how the genetic differences between North and South Wales are playing out with AncestryDNA and also whether your South Wales Border in cautious mode is linked to your ancestryDNA results.

With mine I think there is consistency between the data from both tests as my 13% "Ireland/Scotland" is very close to LivingDNA's 15.8% North Wales.

Phoebe Watts
09-21-2018, 09:23 PM
Can you remind me again Phoebe, what did LivingDNA assign you for North Wales and South Wales in complete mode? Your case is interesting because you have ancestry from north and south.



Here they are, with some explanation:

%
39.7 North Wales (mainly my mother - but including half my father's 8.7% North Wales)
4.4 North West England (my mother's Flintshire grandfather?)
27.7 South Wales (half my father's 53.6%)
10.0 South Central England (looks like half my father's South Wales borders)

My 50/50 in Ancestry is not unlike results of relatives and DNA relatives from North Wales, so I was guessing that there is a West/ East effect rather than North/ South.

I have also seen the results of a relative from south-west Wales (LDNA upload to My Heritage) that might be relevant. That shows 87% Irish, Scottish, Welsh and 13% North and West European.

JonikW
09-21-2018, 09:33 PM
Here they are, with some explanation:

%
39.7 North Wales (mainly my mother - but including half my father's 8.7% North Wales)
4.4 North West England (my mother's Flintshire grandfather?)
27.7 South Wales (half my father's 53.6%)
10.0 South Central England (looks like half my father's South Wales borders)

My 50/50 in Ancestry is not unlike results of relatives and DNA relatives from North Wales, so I was guessing that there is a West/ East effect rather than North/ South.

I have also seen the results of a relative from south-west Wales (LDNA upload to My Heritage) that might be relevant. That shows 87% Irish, Scottish, Welsh and 13% North and West European.

Have you transferred to MyHeritage? If so, what did you get?

Phoebe Watts
09-21-2018, 09:36 PM
Have you transferred to MyHeritage? If so, what did you get?

I have - I'll post the results when they appear.

JonikW
09-21-2018, 09:39 PM
I have - I'll post the results when they appear.

Thanks. There are some posts here that you might find interesting.

https://anthrogenica.com/showthread.php?15323-MyHeritage-accepting-LivingDNA-uploads!&p=493669#post493669

Phoebe Watts
09-22-2018, 09:03 AM
Have you transferred to MyHeritage? If so, what did you get?

Here they are:
Irish, Scottish, Welsh 87.6%; North and West European 11%; Eastern Europe 1.4%

avalon
09-22-2018, 05:50 PM
Here they are, with some explanation:

%
39.7 North Wales (mainly my mother - but including half my father's 8.7% North Wales)
4.4 North West England (my mother's Flintshire grandfather?)
27.7 South Wales (half my father's 53.6%)
10.0 South Central England (looks like half my father's South Wales borders)

My 50/50 in Ancestry is not unlike results of relatives and DNA relatives from North Wales, so I was guessing that there is a West/ East effect rather than North/ South.

I have also seen the results of a relative from south-west Wales (LDNA upload to My Heritage) that might be relevant. That shows 87% Irish, Scottish, Welsh and 13% North and West European.

Interesting. I think there are general east-west effects within Wales as well as north-south effects but I'd like to see some more Welsh results from LivingDNA and AncestryDNA to be able to draw a better conclusion. A lot of it though will come down to an individual's specific ancestry which could vary, even between people from the same locality within Wales, given how mixed modern populations tend to be anyway.

I am seeing at the moment though a consistency between AncestryDNA and LivingDNA data. My cautious mode with LivingDNA matches very closely my AncestryDNA results - NW England 78.3% + 5.8% unassigned is 84.1% which is very close to the 87% that Ancestry gave me for "England, Wales and NW Europe."

I suspect that in your case, LivingDNA gave you "South Wales Border" ancestry in cautious mode for the same reason that Ancestry gave you 50% "England, Wales, NW Europe."

sktibo
09-22-2018, 06:11 PM
@Avalon

My current theory is that Ancestry did add some Welsh samples to the England Wales and NW Europe category but the Welsh samples would be from the borders as those cluster more closely to England than to the extremities of Wales IIRC. I think I remember Living DNA matching you quite strongly to NW England which we figured out was actually something of a Welsh Borders region - sometimes called "Cheshire" in some studies. Perhaps your results hint that this could be what Ancestry did?

Also, it is great that you got an Ancestry test done! You should pick up the latest Eurogenes Celtic vs Germanic PCA and join us down in that thread too. If you see this Phoebe I'm sure everyone would be very interested to see where you plot on that as well.

Phoebe Watts
09-22-2018, 09:56 PM
Interesting. I think there are general east-west effects within Wales as well as north-south effects but I'd like to see some more Welsh results from LivingDNA and AncestryDNA to be able to draw a better conclusion. A lot of it though will come down to an individual's specific ancestry which could vary, even between people from the same locality within Wales, given how mixed modern populations tend to be anyway.

I am seeing at the moment though a consistency between AncestryDNA and LivingDNA data. My cautious mode with LivingDNA matches very closely my AncestryDNA results - NW England 78.3% + 5.8% unassigned is 84.1% which is very close to the 87% that Ancestry gave me for "England, Wales and NW Europe."

I suspect that in your case, LivingDNA gave you "South Wales Border" ancestry in cautious mode for the same reason that Ancestry gave you 50% "England, Wales, NW Europe."

I'm not sure that I follow. Does that mean that you would expect unmixed testers from the north-west (and south-west?) of Wales to score significantly more than 50% Ireland and Scotland? That isn't what I have seen so far. But it would be interesting to see other results.

rms2
09-22-2018, 10:05 PM
I am sometimes tempted to test with Living DNA, but, even though I am mostly British Isles/Irish, my British Isles/Irish is the product of ancestors from different parts of Britain and Ireland who met and bred here in North America, making me a kind of British/Irish mutt.

That's one of the reasons why I focus more on the y chromosome than on anything else.

JonikW
09-22-2018, 10:25 PM
I am sometimes tempted to test with Living DNA, but, even though I am mostly British Isles/Irish, my British Isles/Irish is the product of ancestors from different parts of Britain and Ireland who met and bred here in North America, making me a kind of British/Irish mutt.

That's one of the reasons why I focus more on the y chromosome than on anything else.

I think you're right. Stick with the facts. That's never a bad policy. My Living DNA turned out quite accurately according to my paper trail but in general I think British populations are so close that separating them is problematic and subject to error.

rms2
09-22-2018, 10:30 PM
I think you're right. Stick with the facts. That's never a bad policy.

The y side is easy to trace, and my closest y-dna matches who can get across the Pond are Welsh. So, I kind of identify as Welsh, or as an American of Welsh descent.

But autosomally my biggest, strongest dna circle at Ancestry is one derived from a Scottish ancestor.

JonikW
09-22-2018, 10:47 PM
The y side is easy to trace, and my closest y-dna matches who can get across the Pond are Welsh. So, I kind of identify as Welsh, or as an American of Welsh descent.

But autosomally my biggest, strongest dna circle at Ancestry is one derived from a Scottish ancestor.

You're an old hand so I guess you've done Lucasz's analysis. That correlated well with my Living DNA but cost only a few bucks. That said, I've just been in contact with a close MyHeritage DNA match who says she has also done Living DNA and I'm hoping the new matching feature will turn up more such matches. I like what Living DNA are trying to do and am glad I supported that.

avalon
09-23-2018, 09:58 AM
@Phoebe/sktibo,

Yes, based on my own results with Ancestry and LivingDNA I think there is a general east-west effect within Wales so "Welsh Borders" type ancestry clusters more with England whereas West Wales is more distinct, which is no surprise given the history. But this is a generalisation though, there will always be variations at an individual level.

In my case, Ancestry gave me 13% "Ireland/Scotland" which is incredibly close to my paper trail 12.5% for Merionethshire (NW Wales) which could just be a coincidence, but this is supported by the figures for my LivingDNA cautious view which back up my Ancestry results.

However, if Phoebe has seen results with people whose ancestry is 100% from NW Wales and who are also getting 50/50 with AncestryDNA then this puts a dent in my theory!! :)

It could be that Ancestry's assignment of Welsh ancestry is a bit more random than I give it credit for - I suspect they don't have that many Welsh samples, might be an area they need to improve as they appear much better for Irish and Scottish ancestry.

By the way sktibo, the Eurogenes Celtic/Germanic looks good, I'm glad that he finally has some Welsh samples, Eurogenes usually lacked this in previous tests.

avalon
09-23-2018, 10:06 AM
I am sometimes tempted to test with Living DNA, but, even though I am mostly British Isles/Irish, my British Isles/Irish is the product of ancestors from different parts of Britain and Ireland who met and bred here in North America, making me a kind of British/Irish mutt.

That's one of the reasons why I focus more on the y chromosome than on anything else.

I have seen the LivingDNA results for a few Americans who got significant amounts from Welsh regions, I think "Robert1" was one such poster, from memory. If you've got Welsh names in your tree then I think there's a good chance LivingDNA would pick up on your Welsh ancestry.

sktibo
09-23-2018, 10:59 AM
@Phoebe/sktibo,

Yes, based on my own results with Ancestry and LivingDNA I think there is a general east-west effect within Wales so "Welsh Borders" type ancestry clusters more with England whereas West Wales is more distinct, which is no surprise given the history. But this is a generalisation though, there will always be variations at an individual level.

In my case, Ancestry gave me 13% "Ireland/Scotland" which is incredibly close to my paper trail 12.5% for Merionethshire (NW Wales) which could just be a coincidence, but this is supported by the figures for my LivingDNA cautious view which back up my Ancestry results.

However, if Phoebe has seen results with people whose ancestry is 100% from NW Wales and who are also getting 50/50 with AncestryDNA then this puts a dent in my theory!! :)

It could be that Ancestry's assignment of Welsh ancestry is a bit more random than I give it credit for - I suspect they don't have that many Welsh samples, might be an area they need to improve as they appear much better for Irish and Scottish ancestry.

By the way sktibo, the Eurogenes Celtic/Germanic looks good, I'm glad that he finally has some Welsh samples, Eurogenes usually lacked this in previous tests.

they added a lot of references so it might not be that sparse. Their Irish/Scottish seems really good with the exception of one set of results from a friend I saw.. they weren't assigned any Ireland/Scotland and half of their ancestry is from Northern Ireland, Lowland Scotland, and Ireland - they even land in the Southern Scotland and Northern Ireland genetic community. Their DNA results always seem to give them a bunch of English, even an English GC, so perhaps they had a major NPE going on in their background. Something I don't want to suggest to them however.... Anyhow, they were assigned almost 70% England/Wales & NW Europe and they don't have any ancestry from those areas that was recorded on paper.

Phoebe Watts
09-24-2018, 10:47 AM
@Phoebe/sktibo,

Yes, based on my own results with Ancestry and LivingDNA I think there is a general east-west effect within Wales so "Welsh Borders" type ancestry clusters more with England whereas West Wales is more distinct, which is no surprise given the history. But this is a generalisation though, there will always be variations at an individual level.

In my case, Ancestry gave me 13% "Ireland/Scotland" which is incredibly close to my paper trail 12.5% for Merionethshire (NW Wales) which could just be a coincidence, but this is supported by the figures for my LivingDNA cautious view which back up my Ancestry results.

However, if Phoebe has seen results with people whose ancestry is 100% from NW Wales and who are also getting 50/50 with AncestryDNA then this puts a dent in my theory!! :)

It could be that Ancestry's assignment of Welsh ancestry is a bit more random than I give it credit for - I suspect they don't have that many Welsh samples, might be an area they need to improve as they appear much better for Irish and Scottish ancestry.

By the way sktibo, the Eurogenes Celtic/Germanic looks good, I'm glad that he finally has some Welsh samples, Eurogenes usually lacked this in previous tests.


I can see many Welsh matches where Ireland & Scotland is first on their list of regions, followed by England, Wales etc. but I can't see the percentages for many testers. A relative with all his well-researched recent ancestry from Anglesey, Caernarvonshire and western Denbighshire scores 53% Ireland & Scotland. I'm sure there must be higher Ireland & Scotland scores for Welsh people too!

I guess the 13% ish in both tests is a coincidence and that Merionethshire ancestry will score a high level of Irish and Scottish on Ancestry; Denbighshire and Flintshire a slightly lower level; North West England lower again.

Just to complicate things: I can see the scores in My Heritage. My LDNA transfer kit scores almost 88% Ireland, Scotland, Wales but I can see some matches that I know are third and fourth cousins in a Carmarthenshire line and who score up to 100%. They may have taken a different test of course.

Perhaps we should take sktibo's advice?

Robert1
09-24-2018, 03:50 PM
I have seen the LivingDNA results for a few Americans who got significant amounts from Welsh regions, I think "Robert1" was one such poster, from memory. If you've got Welsh names in your tree then I think there's a good chance LivingDNA would pick up on your Welsh ancestry.

My Mother got quite high Welsh related in Living DNA cautious mode, about 45%. The lion's share is South Wales (she has a grandmother from Glamorgan), her North Wales is around 5% in Complete mode. In the AncestryDNA update she gets 68% Ireland&Scotland and 32% British, Wales & NW European. (Her Welsh has to be in Ireland&Scotland no matter how Ancestry labels it!) Living DNA assigns her 98.2% British Isles and 1.8% Scandinavian. She has a lot of Scots from Argyll, about 25% (she has a grandmother from there, too!).

My Welsh related in Living DNA cautious mode is only about 12% but my known ancestry is much higher, about 20%. Again most is South Wales. In the AncestryDNA update I get 58% Ireland&Scotland and 42% British, Wales & NW European. Living DNA assigns me 98.7% British Isles and 1.3% Scandinavian. I also have a lot of Scots from the Highlands and Argyll, about 28%.

My paternal aunt gets less Welsh related in Living DNA cautious mode, about 8%. In the AncestryDNA update she gets 20% Ireland&Scotland and 80% British, Wales & NW European. Living DNA assigns her 90% British Isles, 5% Europe South and 5% Scandinavian. She has a lot of Highland Scots, about 25%.

These figures are off the top of my head and as we have company staying with us for a week I may not be able to get to actual figures but this is close for now. Like you, Avalon, I do see a strong correlation between Living DNA and the new AncestryDNA estimates in our family tests.

Jessie
09-24-2018, 04:10 PM
I don't really see a correlation between Living DNA and Ancestry in regards to Irish people. Irish results are all over the place on Living DNA but Ancestry is giving most Irish 100% Ireland and Scotland.

sktibo
09-24-2018, 04:25 PM
FWIW Living DNA and Ancestry DNA don't correlate in my case

JMcB
09-24-2018, 05:29 PM
In my case, I would say there is now a correlation between Ancestry & LivingDNA, when it comes to my broad numbers. With perhaps, a slight edge to LivingDNA. However, when it comes to the Irish numbers, Ancestry has now moved into the front and seems more in line with what I would expect. LivingDNA has long had a problem with their Irish results because of their small sample size. Which hopefully, they’re in the process of finally fixing. They under reported mine by approximately 11%. Which is not a major offense by any means and to be honest, if I hadn’t seen how off their numbers were for people who were actually Irish, I might have just chalked it up to recombination. But I did see those errors and Ancestry’s new results have now reconfirmed my initial impression, that mine were off, too.

Here are Ancestry’s results:

England, Wales & Northwestern Europe 49%
Ireland & Scotland 41%
Germanic Europe 3%
Norway 3%
France 2%
Italy 2%

My estimates and LivingDNA’s results are below.

In my opinion, LivingDNA has taken my Irish numbers and shifted them into my English results.

avalon
09-24-2018, 08:08 PM
I can see many Welsh matches where Ireland & Scotland is first on their list of regions, followed by England, Wales etc. but I can't see the percentages for many testers. A relative with all his well-researched recent ancestry from Anglesey, Caernarvonshire and western Denbighshire scores 53% Ireland & Scotland. I'm sure there must be higher Ireland & Scotland scores for Welsh people too!

I guess the 13% ish in both tests is a coincidence and that Merionethshire ancestry will score a high level of Irish and Scottish on Ancestry; Denbighshire and Flintshire a slightly lower level; North West England lower again.

Just to complicate things: I can see the scores in My Heritage. My LDNA transfer kit scores almost 88% Ireland, Scotland, Wales but I can see some matches that I know are third and fourth cousins in a Carmarthenshire line and who score up to 100%. They may have taken a different test of course.

Perhaps we should take sktibo's advice?

It would be useful to see more AncestryDNA results for Welsh people to see if there is a general pattern. I think it is clear that Welsh Borders type ancestry is falling in to the "England, Wales, NW Euro" category (certainly in my case it is) and logically West Wales ancestry should be more Irish like but maybe the AncestryDNA test is not as sophisticated as I think it is and I am reading too much into it.

We could do with more information about the reference panel used by Ancestry, but unfortunately the white paper gives no information about Welsh samples and location.

Going back to LivingDNA, from what I've seen, Welsh results are reasonably good and my cautious mode makes sense in terms of the regions collected together. What was your grouping for Cautious mode again?

Phoebe Watts
09-24-2018, 10:50 PM
@ sktibo Here are the LDNA Cautious and Standard results.

I think there is an eastward pull in Cautious that isn't there in Standard. It maybe the same as the Dutch in K36 and the North and West European from the My Heritage upload.

sktibo
09-26-2018, 06:41 AM
I can see many Welsh matches where Ireland & Scotland is first on their list of regions, followed by England, Wales etc. but I can't see the percentages for many testers. A relative with all his well-researched recent ancestry from Anglesey, Caernarvonshire and western Denbighshire scores 53% Ireland & Scotland. I'm sure there must be higher Ireland & Scotland scores for Welsh people too!

I guess the 13% ish in both tests is a coincidence and that Merionethshire ancestry will score a high level of Irish and Scottish on Ancestry; Denbighshire and Flintshire a slightly lower level; North West England lower again.

Just to complicate things: I can see the scores in My Heritage. My LDNA transfer kit scores almost 88% Ireland, Scotland, Wales but I can see some matches that I know are third and fourth cousins in a Carmarthenshire line and who score up to 100%. They may have taken a different test of course.

Perhaps we should take sktibo's advice?

MyHeritage strikes me as supremely weird. I noticed three of my Scottish DNA relatives (not canadian or american, living in Scotland, ancestors from Scotland) get mostly English and a lot of Scandinavian, with some Native American, IIRC one has north African... One has 6% Ireland/Scotland. The other two get none. I can't really take it seriously as a point of comparison any more. Was my advice to get the Celtic/Germanic PCA?

Phoebe Watts
09-26-2018, 12:16 PM
MyHeritage strikes me as supremely weird. I noticed three of my Scottish DNA relatives (not canadian or american, living in Scotland, ancestors from Scotland) get mostly English and a lot of Scandinavian, with some Native American, IIRC one has north African... One has 6% Ireland/Scotland. The other two get none. I can't really take it seriously as a point of comparison any more. Was my advice to get the Celtic/Germanic PCA?

Thanks - I'm seeing some strange mixes for MyHeritage now too!

Yes Celtic/Germanic PCA. I have posted my results on that thread. It looks like Dutch or germanic again - it would be great to have an explanation

EDIT: I had meant to say "a bit" Dutch or Germanic

sktibo
09-26-2018, 12:48 PM
Thanks - I'm seeing some strange mixes for MyHeritage now too!

Yes Celtic/Germanic PCA. I have posted my results on that thread. It looks like Dutch or germanic again - it would be great to have an explanation

That's really weird it shouldn't be plotting there for you. It worked beautifully well for so many others
Edit: I checked your position it appears although not super Celtic you do fall within the English and the Scottish clusters, and you border on the Irish one too.

avalon
09-26-2018, 07:27 PM
@ sktibo Here are the LDNA Cautious and Standard results.

I think there is an eastward pull in Cautious that isn't there in Standard. It maybe the same as the Dutch in K36 and the North and West European from the My Heritage upload.

On the face of it your South Wales Border looks odd given you have no ancestry from the Welsh Borders.

However, I think I mentioned this ages ago but if you look at the POBI cluster maps that show all clusters down to K=52 then at the final level of analysis, there were about 5 clusters in Pembrokeshire and interestingly one of these clusters also had genetic matches in the Welsh Borders region and a couple of matches in SE Wales, so it looks to me as if there is a type of ancestry in Pembrokeshire that is similar to Welsh Borders ancestry. This might explain your cautious mode shift from standard. I notice that you mentioned an "Eastern" shift on several DNA tests so maybe there is something to this.

As for the explanation , one possibility might be that if you have ancestry from SW Wales then "Little England beyond Wales" might be it. The whole Anglo-Norman medieval era in Wales 1080 to 1400 is a such an important period in the history there must have been some noteworthy population changes. You could argue that the next big population change in Wales doesn't happen until the industrial era post 1700.

The medieval documentary evidence though is quite scarce, we know about the Flemings in Pembrokeshire for instance, but the historian RR Davies makes the point that in the castle garrison towns established all over Wales by the Normans there would have been English settlers drawn from their Lord's English estates but I think we will need a detailed DNA study of Wales before we can assess the genetic impact of the medieval era.

msmarjoribanks
09-26-2018, 08:01 PM
LivingDNA seemed to pick up well on my Welsh -- my known Welsh is about an eighth, half of that from North Wales, and it pegged that. I suspect I have more Welsh ancestry, but that's too far back or too hard to separate from the English (Midlands/border ancestry).

I'm hesitant to say it's really accurate in this, though, until I see what they do when the Irish and German projects are added, since right now they completely missed both my German (10-15%), and I'm assuming that's distorting some of my English results, and my Scots-Irish. Not only do I not get any Irish, but I don't get Scottish at all either.

I found the test very interesting anyway, and am really in it in the hope they start matching. One branch of my dad's family came to the US from England in 1870, and that ancestor married a woman whose parents were both born in Wales, and those are the lines on which I get comparatively few matches at Ancestry -- and the matches I get are either in the US or Canada. So longshot, but a pool that may be more UK based seems appealing.

msmarjoribanks
09-26-2018, 08:03 PM
MyHeritage strikes me as supremely weird. I noticed three of my Scottish DNA relatives (not canadian or american, living in Scotland, ancestors from Scotland) get mostly English and a lot of Scandinavian, with some Native American, IIRC one has north African... One has 6% Ireland/Scotland. The other two get none. I can't really take it seriously as a point of comparison any more. Was my advice to get the Celtic/Germanic PCA?

MyHeritage strikes me as weird with no real pattern, or a pattern depending on when you upload the test.

Opposite of your relatives, it overstates my mom's Irish/Scottish, and gives her no English at all (she's about equal portions of both).

JonikW
09-26-2018, 09:01 PM
That's really weird it shouldn't be plotting there for you. It worked beautifully well for so many others
Edit: I checked your position it appears although not super Celtic you do fall within the English and the Scottish clusters, and you border on the Irish one too.

I agree with you sktibo (and msmarjoribanks) about MyHeritage. It throws up some weird stuff in my opinion. I recommend Living DNA, Lucasz and now in particular the Celtic/Germanic test for those with Isles/NW Europe ancestry. The latter has just given me plenty of food for thought. Where MyHeritage excelled for me was in a bunch of close matches to my late mother. I'm talking people who actually knew her father. Amazing.

msmarjoribanks
09-26-2018, 09:11 PM
Yeah, it's worth it for the matches, although so far I don't have as many good matches on MyHeritage as FTDNA and Ancestry.

I need to figure out how to do the Lucasz and Celtic/Germanic tests.

JonikW
09-26-2018, 09:40 PM
Yeah, it's worth it for the matches, although so far I don't have as many good matches on MyHeritage as FTDNA and Ancestry.

I need to figure out how to do the Lucasz and Celtic/Germanic tests.

Please do. Even I managed it. :) I'd love to see your results.

Phoebe Watts
09-26-2018, 10:19 PM
On the face of it your South Wales Border looks odd given you have no ancestry from the Welsh Borders.

However, I think I mentioned this ages ago but if you look at the POBI cluster maps that show all clusters down to K=52 then at the final level of analysis, there were about 5 clusters in Pembrokeshire and interestingly one of these clusters also had genetic matches in the Welsh Borders region and a couple of matches in SE Wales, so it looks to me as if there is a type of ancestry in Pembrokeshire that is similar to Welsh Borders ancestry. This might explain your cautious mode shift from standard. I notice that you mentioned an "Eastern" shift on several DNA tests so maybe there is something to this.

As for the explanation , one possibility might be that if you have ancestry from SW Wales then "Little England beyond Wales" might be it. The whole Anglo-Norman medieval era in Wales 1080 to 1400 is a such an important period in the history there must have been some noteworthy population changes. You could argue that the next big population change in Wales doesn't happen until the industrial era post 1700.

The medieval documentary evidence though is quite scarce, we know about the Flemings in Pembrokeshire for instance, but the historian RR Davies makes the point that in the castle garrison towns established all over Wales by the Normans there would have been English settlers drawn from their Lord's English estates but I think we will need a detailed DNA study of Wales before we can assess the genetic impact of the medieval era.

Thanks Avalon, those are useful points.

I do have ancestry at and below the Landsker in Pembrokeshire and although they seem to Welsh speaking in the late 1700s, they could well have been of settler stock. You'll know that the south Wales coast was part of the March and was settled by Normans early. Some of the locations in my family tree are close to Norman castles at Roch, Llansteffan and Laugharne.

Having said that, the south Wales is on my father's side and he doesn't seem to be affected by the eastern shift. I'm wondering whether the Flintshire/ north-west England and Anglo-Norman elements from my mother's side are enough to shift from South Wales to South Wales Borders. I can trace a small part of my Anglesey grandfather's family back to Anglo-Norman nobilty brought in after 1282 - they married Welsh wives to get hold of their family land. There were some mixed marriages after the battle of Bosworth too. It is just that the numbers are small and that it seems too long ago. But there could be something that I haven't been able to trace.

I'll have to get more relatives to test. It would be good to see more Welsh samples in the PCA too.

JonikW
09-26-2018, 11:13 PM
Thanks Avalon, those are useful points.

I do have ancestry at and below the Landsker in Pembrokeshire and although they seem to Welsh speaking in the late 1700s, they could well have been of settler stock. You'll know that the south Wales coast was part of the March and was settled by Normans early. Some of the locations in my family tree are close to Norman castles at Roch, Llansteffan and Laugharne.

Having said that, the south Wales is on my father's side and he doesn't seem to be affected by the eastern shift. I'm wondering whether the Flintshire/ north-west England and Anglo-Norman elements from my mother's side are enough to shift from South Wales to South Wales Borders. I can trace a small part of my Anglesey grandfather's family back to Anglo-Norman nobilty brought in after 1282 - they married Welsh wives to get hold of their family land. There were some mixed marriages after the battle of Bosworth too. It is just that the numbers are small and that it seems too long ago. But there could be something that I haven't been able to trace.

I'll have to get more relatives to test. It would be good to see more Welsh samples in the PCA too.

Hi Phoebe, I was particularly interested in your Celtic vs Germanic plot today in light of your wider signals from various tests. I'm not surprised about my own slightly Germanic shifted result because although I've got strong Welsh/Celtic signals on some platforms I know my ancestry in recent generations includes both Welsh and English (I also have Scottish on at least two more lines and Irish on a more distant one). We don't seem to have many fully Welsh testers here and in light of Avalon's excellent post do you have a simple list of places associated with your 32 ggg grandparents, whether place of birth, residence or death? I'd be interested to see it if so. As far as I remember you're a mixture of North and South and a list might help to build a picture as to whether East or West could be responsible for any Germanic signal.

Phoebe Watts
09-27-2018, 09:24 AM
Hi Phoebe, I was particularly interested in your Celtic vs Germanic plot today in light of your wider signals from various tests. I'm not surprised about my own slightly Germanic shifted result because although I've got strong Welsh/Celtic signals on some platforms I know my ancestry in recent generations includes both Welsh and English (I also have Scottish on at least two more lines and Irish on a more distant one). We don't seem to have many fully Welsh testers here and in light of Avalon's excellent post do you have a simple list of places associated with your 32 ggg grandparents, whether place of birth, residence or death? I'd be interested to see it if so. As far as I remember you're a mixture of North and South and a list might help to build a picture as to whether East or West could be responsible for any Germanic signal.

I hope this will do - places of birth of the ggg is the same as the place located with their fathers on this thread:

https://anthrogenica.com/showthread.php?12269-What-did-your-4-x-great-grandfathers-(Generation-7)-do-for-a-living&p=468558&viewfull=1#post468558

JonikW
09-27-2018, 10:11 AM
I hope this will do - places of birth of the ggg is the same as the place located with their fathers on this thread:

https://anthrogenica.com/showthread.php?12269-What-did-your-4-x-great-grandfathers-(Generation-7)-do-for-a-living&p=468558&viewfull=1#post468558

What a wonderful list. So predominantly West and North with little in the South and nothing in the South East. No obvious Germanic clues there then unless as you say it could be English/Flemings in Pembrokeshire. Perhaps the Norse and English had some impact on Anglesey, as a few people have theorised.

Solothurn
09-27-2018, 01:31 PM
https://www.yourdnaportal.com/ have the Lucasz calc. If it is the correct one that is :(





I need to figure out how to do the Lucasz and Celtic/Germanic tests.

Nqp15hhu
09-27-2018, 08:44 PM
Would love to do this but canít because iím from Northern Ireland.

JonikW
09-27-2018, 08:54 PM
Would love to do this but can’t because i’m from Northern Ireland.

Can't do what nqp? Is there something you can't access from there?

Nqp15hhu
09-27-2018, 09:40 PM
The results are heavily skewed to GB.

JonikW
09-27-2018, 10:05 PM
The results are heavily skewed to GB.

You're right there. We need the update asap. Then it might be worth you giving it a shot.

avalon
09-29-2018, 04:06 PM
Thanks Avalon, those are useful points.

I do have ancestry at and below the Landsker in Pembrokeshire and although they seem to Welsh speaking in the late 1700s, they could well have been of settler stock. You'll know that the south Wales coast was part of the March and was settled by Normans early. Some of the locations in my family tree are close to Norman castles at Roch, Llansteffan and Laugharne.

Having said that, the south Wales is on my father's side and he doesn't seem to be affected by the eastern shift. I'm wondering whether the Flintshire/ north-west England and Anglo-Norman elements from my mother's side are enough to shift from South Wales to South Wales Borders. I can trace a small part of my Anglesey grandfather's family back to Anglo-Norman nobilty brought in after 1282 - they married Welsh wives to get hold of their family land. There were some mixed marriages after the battle of Bosworth too. It is just that the numbers are small and that it seems too long ago. But there could be something that I haven't been able to trace.

I'll have to get more relatives to test. It would be good to see more Welsh samples in the PCA too.

The problems is, as you know, is that Wales is desperately in need of a good genetic survey that includes ancientDNA from all periods of history/pre-history to compare to modern populations. My own guess is that if we are looking for an explanation for a "Germanic" shift in the modern Welsh then the first place to look is the Medieval Anglo-Norman era 1100-1300 roughly, because the history and the archaeology suggests the Anglo-Saxons had minimal or no impact on Wales. You've then got the period 1400 to 1700 but I don't think there are any major population changes again until the industrial era post-1700.

For North Wales, and perhaps for other parts of Wales, I see the Anglo-Norman conquest and English settlement of Wales as patchy. For instance, in Edward I's newly established castle towns of North Wales such as Beaumaris, Conwy, Harlech, Denbigh, Ruthin, it looks like they were quite successful "plantation boroughs" but others like Criccieth were not. And it looks like much of the North Wales countryside, the hills and the mountains were left alone by English settlers. So you could have a situation where the impact of the Edwardian settlement of North Wales differed between coastal towns and the hinterlands such as in Snowdonia.

Of course, this was all along time ago, so in the centuries following 1282, populations in North Wales may have become more mixed anyway. I do think the "Eastern" shift in Flintshire may have occurred at around this time as we know of settlement in the Vale of Clwyd in the 1280s which is very close to where much of my ancestry is from. Of course, my Flintshire ancestors had Welsh surnames and spoke Welsh, but I guess it's possible that Medieval English settlers in Wales did eventually integrate into Welsh communities and take Welsh names centuries later.

Phoebe Watts
09-29-2018, 05:47 PM
The problems is, as you know, is that Wales is desperately in need of a good genetic survey that includes ancientDNA from all periods of history/pre-history to compare to modern populations. My own guess is that if we are looking for an explanation for a "Germanic" shift in the modern Welsh then the first place to look is the Medieval Anglo-Norman era 1100-1300 roughly, because the history and the archaeology suggests the Anglo-Saxons had minimal or no impact on Wales. You've then got the period 1400 to 1700 but I don't think there are any major population changes again until the industrial era post-1700.

For North Wales, and perhaps for other parts of Wales, I see the Anglo-Norman conquest and English settlement of Wales as patchy. For instance, in Edward I's newly established castle towns of North Wales such as Beaumaris, Conwy, Harlech, Denbigh, Ruthin, it looks like they were quite successful "plantation boroughs" but others like Criccieth were not. And it looks like much of the North Wales countryside, the hills and the mountains were left alone by English settlers. So you could have a situation where the impact of the Edwardian settlement of North Wales differed between coastal towns and the hinterlands such as in Snowdonia.

Of course, this was all along time ago, so in the centuries following 1282, populations in North Wales may have become more mixed anyway. I do think the "Eastern" shift in Flintshire may have occurred at around this time as we know of settlement in the Vale of Clwyd in the 1280s which is very close to where much of my ancestry is from. Of course, my Flintshire ancestors had Welsh surnames and spoke Welsh, but I guess it's possible that Medieval English settlers in Wales did eventually integrate into Welsh communities and take Welsh names centuries later.

It is surprising how quickly new settlers could be assimilated given the right circumstances.

Have you seen the discussion over on the Eurogenes Celtic vs Germanic PCA thread?

It is interesting PCA - I had good feedback here:

https://anthrogenica.com/showthread.php?15388-New-Eurogenes-Celtic-vs-Germanic-PCA&p=496365&viewfull=1#post496365
https://anthrogenica.com/showthread.php?15388-New-Eurogenes-Celtic-vs-Germanic-PCA&p=496400&viewfull=1#post496400

This is Agamemnon's take on my Eurogenes PCA: "Firmly British, there does seem to be a noticeable pull towards Germanic populations, but you're still very close to the Welsh."

https://anthrogenica.com/showthread.php?15388-New-Eurogenes-Celtic-vs-Germanic-PCA&p=497260&viewfull=1#post497260
is interesting too.

jshook
09-30-2018, 12:32 AM
I have seen the LivingDNA results for a few Americans who got significant amounts from Welsh regions, I think "Robert1" was one such poster, from memory. If you've got Welsh names in your tree then I think there's a good chance LivingDNA would pick up on your Welsh ancestry.

For whatever reason, I got about 8.5% South Wales Border (which they describe as Shropshire, Herefordshire, Monmouthshire, Worcestershire, Powys and Gwent), which was my 3rd largest region. But every Welsh ancestor I've found in my family tree is from the northern part of Wales (Conwy and Denbighshire primarily), which just didn't show up at all. It's certainly possible that it just washed out over time and that 8.5% is 100% English that they're picking up. But I'm skeptical. I also get a phantom 2.9% result from the neighboring Northwest England region when I have no known ancestors from that area. Make of that what you will.

Phoebe Watts
09-30-2018, 09:07 AM
For whatever reason, I got about 8.5% South Wales Border (which they describe as Shropshire, Herefordshire, Monmouthshire, Worcestershire, Powys and Gwent), which was my 3rd largest region. But every Welsh ancestor I've found in my family tree is from the northern part of Wales (Conwy and Denbighshire primarily), which just didn't show up at all. It's certainly possible that it just washed out over time and that 8.5% is 100% English that they're picking up. But I'm skeptical. I also get a phantom 2.9% result from the neighboring Northwest England region when I have no known ancestors from that area. Make of that what you will.


That's interesting. The western part of north Wales seems to be quite easily identified by LDNA as North Wales. A few of us with ancestry in Denbighshire and Flintshire have said it is reported as Northwest England, yours might be the same.

avalon
09-30-2018, 10:03 AM
It is surprising how quickly new settlers could be assimilated given the right circumstances.

Have you seen the discussion over on the Eurogenes Celtic vs Germanic PCA thread?

It is interesting PCA - I had good feedback here:

https://anthrogenica.com/showthread.php?15388-New-Eurogenes-Celtic-vs-Germanic-PCA&p=496365&viewfull=1#post496365
https://anthrogenica.com/showthread.php?15388-New-Eurogenes-Celtic-vs-Germanic-PCA&p=496400&viewfull=1#post496400

This is Agamemnon's take on my Eurogenes PCA: "Firmly British, there does seem to be a noticeable pull towards Germanic populations, but you're still very close to the Welsh."

https://anthrogenica.com/showthread.php?15388-New-Eurogenes-Celtic-vs-Germanic-PCA&p=497260&viewfull=1#post497260
is interesting too.

Yes, I've had a look at that. It does look quite good, I may do that test myself as it is quite cheap.

I think the PCA is giving us a good overview of ancient v modern samples but I still think we are several years away from getting a detailed and comprehensive genetic history of Britain as I think the data at this point is still limited.

avalon
09-30-2018, 10:31 AM
For whatever reason, I got about 8.5% South Wales Border (which they describe as Shropshire, Herefordshire, Monmouthshire, Worcestershire, Powys and Gwent), which was my 3rd largest region. But every Welsh ancestor I've found in my family tree is from the northern part of Wales (Conwy and Denbighshire primarily), which just didn't show up at all. It's certainly possible that it just washed out over time and that 8.5% is 100% English that they're picking up. But I'm skeptical. I also get a phantom 2.9% result from the neighboring Northwest England region when I have no known ancestors from that area. Make of that what you will.

Just to follow up on what Phoebe said, I also got 5.8% South Wales Border from LivingDNA despite having no known ancestry from that region. I have quite a bit of ancestry from Flintshire and Eastern Denbighshire and it looks like this was assigned to NW England by LivingDNA so it looks your Welsh ancestry has been assigned to nearby regions, like mine was.

Essentially the problem is two fold - firstly the original POBI project didn't collect many samples from NE Wales so we are probably lacking data for that region. Secondly, it is a labelling issue. POBI originally had a "Welsh Borders" cluster which was primarily based in Herefordshire/Forest of Dean and in Cheshire but when it came to the LivingDNA test they split the cluster in to South Wales Border and NW England. I think it's highly likely that these 2 regions are very similar genetically and that there is going to be a lot of genetic overlap between the West of England and with Eastern Wales.

msmarjoribanks
10-01-2018, 04:51 PM
For whatever reason, I got about 8.5% South Wales Border (which they describe as Shropshire, Herefordshire, Monmouthshire, Worcestershire, Powys and Gwent), which was my 3rd largest region. But every Welsh ancestor I've found in my family tree is from the northern part of Wales (Conwy and Denbighshire primarily), which just didn't show up at all. It's certainly possible that it just washed out over time and that 8.5% is 100% English that they're picking up. But I'm skeptical. I also get a phantom 2.9% result from the neighboring Northwest England region when I have no known ancestors from that area. Make of that what you will.

My North Wales shows up, but I get lower borders region than expected even though I know I have ancestors from that region (Shropshire near the border) and am assuming that could be part of what my NW England refers to. But until they get my Ulster figured out (as it's just missing) it's hard to say what's what. Some of my Ulster ancestors have English or possibly English names, but I doubt strongly that they were all just English.

jshook
10-01-2018, 05:18 PM
That's interesting. The western part of north Wales seems to be quite easily identified by LDNA as North Wales. A few of us with ancestry in Denbighshire and Flintshire have said it is reported as Northwest England, yours might be the same.

I should also add that that Northwest England result becomes "Unassigned Great Britain" when you switch from Complete to Standard mode (as does my 4.9% Southeast England result.) It would appear that they're definitely not sure it's actually Northwest England.

msmarjoribanks
10-02-2018, 01:09 AM
I should also add that that Northwest England result becomes "Unassigned Great Britain" when you switch from Complete to Standard mode (as does my 4.9% Southeast England result.) It would appear that they're definitely not sure it's actually Northwest England.

Worth doing the comparison.

For me complete and standard modes are basically identical. In those, my English is:

13.1% SE England
8.2% South England
7.9% East Anglia
5.3% Cornwall
4.2% Devon
2.3% South Central England
(so somewhat southern biased, which I think is not unlikely in that I know a number of my 1600s ancestors were from Southern England and my dad's 1870 ancestor was half Essex/half Shropshire, and I have specific known ancestors from Suffolk and Devon).

In addition, I have:

10.1% NW England
4.6% South Yorkshire
2.9% Cumbria
2.6% North Yorkshire
1.7% Central England

For Scotland/Ireland:

1.3% Orkney and Shetland Islands (but this could be related to my actual Scandinavian)

For Wales:

6.2% North Wales (I'd put it at half of my known 1/8 Welsh, so looks good)
3.8% South Wales (another 6.5% should be Montgomeryshire, so not sure where that would go, not really South).
2.5% South Wales Border (as noted I have a decent amount of Shropshire, as well as the Montgomeryshire -- I also have likely but unknown Welsh from ancestors with likely Welsh names who settled in a part of Ohio (Gallia/Jackson counties) known for "Calvinistic" Welsh immigrants in the late 1700s, early 1800s).

Rest not British Isles, according to LivingDNA.

When I switch to Cautious, I get:

27.9% South England related ancestry
17.9% NW England related ancestry (higher)

7.9% East Anglia (same)
6.2% North Wales (same)
5.3% Cornwall (same)
3.8% South Wales (same)
1.3% Orkney-related ancestry (same)

5.6% GB and Ireland unassigned

And then:

12.2% NW Europe related ancestry unassigned (too low)
1.8% Sardinian related ancestry (maybe, beats me)
3.7% Europe unassigned

3.7% NW Caucasus related unassigned (beats me)
1.5% World, unassigned

So interestingly my main English/Welsh categories change very little.

jshook
10-02-2018, 06:02 AM
Worth doing the comparison.

For me complete and standard modes are basically identical. In those, my English is:

13.1% SE England
8.2% South England
7.9% East Anglia
5.3% Cornwall
4.2% Devon
2.3% South Central England
(so somewhat southern biased, which I think is not unlikely in that I know a number of my 1600s ancestors were from Southern England and my dad's 1870 ancestor was half Essex/half Shropshire, and I have specific known ancestors from Suffolk and Devon).

In addition, I have:

10.1% NW England
4.6% South Yorkshire
2.9% Cumbria
2.6% North Yorkshire
1.7% Central England

For Scotland/Ireland:

1.3% Orkney and Shetland Islands (but this could be related to my actual Scandinavian)

For Wales:

6.2% North Wales (I'd put it at half of my known 1/8 Welsh, so looks good)
3.8% South Wales (another 6.5% should be Montgomeryshire, so not sure where that would go, not really South).
2.5% South Wales Border (as noted I have a decent amount of Shropshire, as well as the Montgomeryshire -- I also have likely but unknown Welsh from ancestors with likely Welsh names who settled in a part of Ohio (Gallia/Jackson counties) known for "Calvinistic" Welsh immigrants in the late 1700s, early 1800s).

Rest not British Isles, according to LivingDNA.

When I switch to Cautious, I get:

27.9% South England related ancestry
17.9% NW England related ancestry (higher)

7.9% East Anglia (same)
6.2% North Wales (same)
5.3% Cornwall (same)
3.8% South Wales (same)
1.3% Orkney-related ancestry (same)

5.6% GB and Ireland unassigned

And then:

12.2% NW Europe related ancestry unassigned (too low)
1.8% Sardinian related ancestry (maybe, beats me)
3.7% Europe unassigned

3.7% NW Caucasus related unassigned (beats me)
1.5% World, unassigned

So interestingly my main English/Welsh categories change very little.

Yeah, I also don't want to make it sound like I'm unhappy with LivingDNA at all. Trying to breakdown any nation this finely is bound to be somewhat inexact. Their Top 2 groupings for me in Cautious mode (27.9% South Central England related and 18.7% North Yorkshire related) correspond extremely well with my known paper trail even if one can quibble about the exact percentages. The main issues for me have been:

1. I'm pretty certain both my Southeast England (4.9%) and South England (6.2%) are reflective of my German ancestry, not English. For comparison, on 23andMe I'm now 14.7% German/French with the latest update. That's about what it should be. I got zero Germanic on LivingDNA and I don't have any known ancestors that match either the SE or S English regions.
2. My Scottish is extremely understated unless I just didn't inherit a lot of Scottish DNA from relatively recent ancestors. Again, I'm skeptical of that.
3. The missing North Welsh ancestry mentioned earlier.

I also don't really grasp how the Cautious groupings work. My top two make total sense. And since Northwest Scotland at 6.4% was my only Scottish region, I get that it'd be a standalone. But my result lists Cumbria as both part of the North Yorkshire Related group AND also it's own 4.8% standalone result. My husband has similar oddities (getting 4.7% North Wales related ancestry when he gets zero North or South Wales in the Standard or Complete Modes and SW Scotland being included in both the NW-Scotland related group and the Cumbria related group.)

msmarjoribanks
10-02-2018, 04:17 PM
I'm not unhappy with it either, just excited to see how improving the German and Irish accuracy will affect mine.

Angriff
10-06-2018, 07:27 AM
I'm not unhappy with it either, just excited to see how improving the German and Irish accuracy will affect mine.

I think the position you describe is where most of us stand. Even the current LivingDNA results show more promise than either past or updated Ancestry or 23andMe, so it is only logical.... really all that remains is upside especially given the contrast between the current sample and those two groups from the two vital directions...

Angriff
10-06-2018, 07:27 AM
I'm not unhappy with it either, just excited to see how improving the German and Irish accuracy will affect mine.

I think the position you describe is where most of us stand. Even the current LivingDNA results show more promise than either past or updated Ancestry or 23andMe, so it is only logical.... really all that remains is upside especially given the contrast between the current sample and those two groups from the two vital directions...

kujira692
10-08-2018, 11:26 PM
Just sharing my own Standard vs. Caution Sub Regional results:

Standard Sub-Region:

Southeast England: 33.9%
East Anglia: 8.9%
South England: 8.4%
Northumbria 8.3%
Ireland: 4.1%
Cornwall: 4%
South Yorkshire: 3.4%
Northwest Scotland 3.4%
Aberdeenshire: 3%
Lincolnshire: 2.7%
Devon: 2.1%
Southwest Scotland and Northern Ireland: 1.7%
Northwest England: 1.5%
North Yorkshire: 1.3%
Great Britain and Ireland (Unassigned): 3.6%

Cautious Sub-Region:

Devon-Related Ancestry: 48.3%
North Yorkshire-Related Ancestry: 15.8%
Ireland-Related Ancestry: 10.7%
East Anglia: 8.9%
Orkney-Related Ancestry: 3%
Great Britain and Ireland (Unassigned): 3.6%

So, right away I'm quite certain that my Southeast England is very inflated from my Southwest German heritage (I do have a Great-Great-Grandfather born in Kent however). East Anglia seems good as I have a Great-Grandmother born in Wormingford, Essex. South England is likely good too since there is ancestry way back from Christchurch, Dorset. The remainder is hard to figure how exactly accurate it is. Also: does it make sense that the sum of the parts that go on to comprise "Devon-Related Ancestry" (namely: Devon, Northumbria(?), and two other regions near Devon that I lack) would go on to comprise nearly half my ancestry when those elements add up to only just over 10% on Standard?

Edit:
Also, part 2: why does the "Ireland-Related Ancestry" Cautious Mode results only amount to 10.7% when it includes, not only Ireland and Western Scotland, but Southeast England too for some reason?

penlanach
11-09-2018, 12:06 PM
This is off the top of my head but it was said that it takes an incredibly close examination to break the large red SE English cluster into separate regions. While some Living DNA test results were able to successfully split the ancestry of some into regions within it for others it was not able to. I personally think the scale is too fine. For all we know someone from a place like Hampshire could have more Germanic ancestry than someone from East Anglia due to later migrations from the continent like France or Belgium. Norfolk's collection of aDNA results reveals a very French-like autosomal picture and for an East Anglian clusters very far from the Germanic types. It's difficult to say with certainty at the time being.

Edit: This youtube video shows and explains how the POBI samples break down into their clusters:
https://www.youtube.com/watch?v=6ANNHMzmxlI
Starts around 16:40
He says the large red cluster never splits even at 53 clusters. I think I remember hearing somewhere that if you go past this level it does eventually split but at that point the differences are incredibly tiny.

The Irish DNA Atlas has one large English cluster (blue on their map instead of read) at K30 but at K48 four English clusters split off from this but they are all very tiny - two have 1 individual and the other two have only 3. So two analyses conclude that it's a pretty darn genetically homogeneous area.

Sorry, not relevant to the LivingDNA discussion here, but I'm just quite astonished that in the video he gives the date of the 'Anglo-Saxon' contribution to the population as 858CE! Garrett says this makes some sense as the intermixing would not occur upon initial arrival (from 400CE onwards), but would occur after years of settlement. But a whole 300-400 years of little to no intermixing between Britons and Germanic settlers? This flies in the face of the emergent archaeological consensus.

I'm almost tempted to say that this study shows that the Germanic contributions to British DNA cannot be definitively pinpointed to a specific period of mass migration, and that a trickle of migrations occurred throughout the Isles history simply because of proximity; with the post-Roman/early medieval 'migration period' Germanic settlement of England being just one of many.

Capitalis
11-10-2018, 02:27 PM
The recent Alice Roberts BBC show had someone from the POBI study mention that the reason the big red cluster exists isn't because it's the Anglo-Saxon cluster but because this is where the Romans laid the most roads, meaning it was easier for people to move around and intermingle than in less Roman-influenced parts of England.

Also, I found a rather large document (which I took to be a pre-thesis) when searching the internet for studies on the influence of the Franks in England, and in summary, yes there is evidence from Kent to Wiltshire.

Stephen1986
11-11-2018, 11:54 AM
Standard results on LivingDNA -

Europe 98.8%

-Great Britain and Ireland 96.2%
--Northwest England 36.1%
--South England 20.6%
--Cumbria 9.7%
--Southeast England 5.5%
--Southwest Scotland and Northern Ireland 3.3%
--Aberdeenshire 2.5%
--North Wales 2.2%
--Northumbria 2.1%
--Orkney and Shetland Islands 1.2%
--South Wales Border 1.2%

--Great Britain and Ireland unassigned 11.9%
--Europe unassigned 2.6%
--World unassigned 1.2%

Cautious results on LivingDNA -

Europe 98.8%

-Great Britain and Ireland 96.2%
--North Wales related ancestry 39.5%
--South England related ancestry 26%
--Cumbria related ancestry 15.1%
--Orkney related ancestry 3.7%

--Great Britain and Ireland unassigned 11.9%
--Europe unassigned 2.6%
--World unassigned 1.2%

Here's a map of my known ancestry (I don't know two of my maternal great great grandfathers so 1/8th of my ancestry is unknown, although they're both probably from Lancashire). The bulk of my ancestry is from the North West of England, especially Lancashire. I think my LivingDNA results are mostly quite accurate.

27058