Journals of a Time Traveller

I received my son Edward's Living DNA results yesterday, 26 days before the deadline (albeit second sample as first failed. Edward was born severely disabled with severe development delay and doesn't always want to give a swab sample). Here I review all of the Living DNA results against what I believe ancestry is based solely on documented genealogical sources. These documented sources are supported only by family history, interviews, thirty years of documentary research - much of it very local, photographs (likenesses), social background (mainly rural working class, many very localised), local social history, and also in my case by DNA cousin matching:



NPEs and genealogical mistakes (particularly over six generations ago) are possible. However, bearing in mind the above factors, I feel that I have a reasonably good documented record to compare DNA-tests-for-ancestry against.

The two tests in the table below, are my son Edward (left), and myself (right). We are both British by nationality and English by ethnicity. We live in East Anglia, home of many of our documented ancestors. I am mainly of East Anglian ancestry, but with some Oxfordshire, Berkshire, Northants, and Swiss lines. Edward's mother by ancestry is half English (Berks, Wilts, East Anglia, Beds, Somerset) and half Irish.

The "actual documented ancestry" are percentages divided into LDNA sub regions, based on Generation 6 (32 x 3rd great grandparents) with some references to Generation 7 when noteworthy.



Discussion.

Living DNA has made great commercial headway through the use of the POBI dataset, that had been acclaimed by the original team, as demonstrating a distinguishable pattern that aligned well with the Anglo-Saxon period British kingdoms. The reference samples were well chosen, with geographically local born grandparents, and a bias to rural testers over urban (that see more mobility). However, LDNA, we know, have made lot's of alterations with their sub regions.

In both the case of Edward's and my documented East Anglian ancestry, the LDNA saw less than 50% of what I'd expect. I believe that some of our EA went elsewhere - Lincs, SE England, Germanic, Scandinavian, etc.

Edward got a big chunk of Lincs and SE England that I just do not think is real.

I got 10% Tuscan that I don't believe. Something could however correlate to a Swiss 3 x great grandparent. But that percentage? I know it is possible. Edward got no Tuscany.

Edward, with his 25% actual Irish ancestry received 98% GB & Ireland. His father, myself, although 97% actual, received only 70% GB & Ireland. I think that rather like with other test companies - Ireland, Scotland, and Wales look more "British" than do the English on tests. Yet Edward only received a mere 2% Irish on the test! That is an even more serious underscore.

Before I tested with Living DNA, I really was starting to lose faith in autosomal DNA testing for general ancestry. When Living DNA launched, my hopes were raised that with the right references, computing power, and chips, that one day, they could be much more accurate and meaningful than they have been up to now.

After Edward's results, I'm starting to lose a bit of faith again. I feel that these tests are good for pin pointing a corner of Europe at best. Beyond that, there may be average PCA plots, but they are far too fuzzy to base "ancestry composition", my origins", or "ethnicity estimate" percentages on with any degree of certainty or accuracy. Fun yes. Useful to build a personal DNA "ancestral population flavour and PCA plot" yes. But that's all. Beyond that is a roll of the dice. Too many testers take their results far too literally. Too many testers also display brand loyalty.

As for haplogroups, Edward's results were disappointingly basic.

Y-DNA L1

mtDNA H (only 4 mutations listed on the csv file).

Davidski published a spreadsheet of his K7 Basal-rich ancient admixture calculator results.  I've harvested some of those test results to compare to my own.

Sorted to three admixtures.

1. Villabruna (component of Western Hunter-Gatherer) sorted:

2. Basal-rich (component of Early Neolithic Farmer) sorted:

3. Ancient North Eurasian (component of Copper Age Steppe) sorted:

Discussion.

1. Villabruna.

I was surprised just how much Villabruna I had in my results, but clearly, this component is found not only in Villabruna, but also in Iceman / Sardinia (Neolithic Farmer) at 53-55%, and in Steppe Yamna (Bronze Age Steppe) at 34%.  Therefore I have inherited substantial Villabruna not perhaps directly, but from early admixtures via Neolithic and Bronze Age settlement of Western Europe.

On Villabruna sorted, my closest neighbours are indeed other British - English, Scottish, and also Dutch and Anglo-Saxon ancient.  I'm correctly positioned.

2. Basal-rich.

This one is an enigma, but it repeats in other calculators.  I receive more Southern European / Neolithic than do most British or indeed North West Europeans. I do have one Swiss 3 x great grandparent, but otherwise, all of my recorded genealogy is South East English.  I have more Basal-rich than do other "British" results in the spreadsheet.  Why is that?  Could have the randomness of genetic recombination given me just a little more of that Swiss line than I should normally expect, or do I have another ancestor from the South within the past several generations, that I'm not yet aware of?  Living DNA suggests Tuscan in their test.

My neighbours in Basal-rich are East French, Dutch, and Swiss.

3. Ancient north Eurasian.

This trades with my Basal-rich results.  I usually have less Steppe than I would normally expect as a Brit or as a North West European.  More Neolithic, less Bronze Age Steppe.  It does suggest some unknown Southern European ancestry.

My ANE neighbours are East French, Swiss, and North Italian.  One ancient British result not too far away is Romano-British 6DT18.  He was a young male (aged 16 to 18 years at death), in a multiple grave with three other men, at Driffield Terrace, York.  He carried Y haplogroup R1b1a2a1a.

In summary, for a Brit, even for an Englishman, I have just about expected percentages of Villabruna and Western Hunter-Gatherer ancestry.  However, something is atypical in the other two founder mixes.  I look more like French or Swiss, with more Basal-rich (Neolithic) and less ANE (Steppe) than the average British or even English.

 

This is the K11 admixture calculator with rarer alleles created by Dilawer Kahn, now available for a small fee as a test on Geneplaza.  I had previously commissioned Dilawer to run my 23andme DNA raw data through the calculator, but this is a nicer presentation.  The test seeks to estimate ancient ancestry admixture using his rarer alleles principle.

My results:

"These were the indigenous populations of Europe that substantially contributed to the genetics of modern Europeans. It is believed that these hunter gatherers arrived in Europe around 45000 years ago from the Near East.".

My Western European Hunter-gatherer admix is 21.7%

Neolithic European

"This population introduced farming to Europe during the Neolithic, and were very likely descended from Neolithic farmers from the Near East. Their genetic signature is best preserved in modern Sardinians and other southern Europeans.".

My Neolithic European admix is 21.7%

Neolithic Anatolian

"These early farmers from Anatolia from about 8000 years ago were the ancestors of the Early European farmers that introduced farming to SE Europe, and replaced the hunter-gatherer cultures that lived there.".

My Neolithic Anatolian admix is 16.4%

Andronova-Srubnaya

"The Andronovo culture is a collection of similar local Bronze Age cultures that flourished around 3000-4000 years ago in western Siberia and the west Asiatic steppe. This culture overlapped with the Srubna culture in the Volga-Ural region of Russia.".

My Andronova-Srubnaya admix is 14.6%

Yamnaya-Poltavka

"The Yamna culture (also known as the Pit Grave culture), was an early Bronze Age culture from the Pontic Eurasian steppe from around 5000 years ago. The Yamna culture is identified with the late Proto-Indo-Europeans, and is the strongest candidate for the homeland of the Proto-Indo-European language.

My Yamnaya-Poltavka admix is 12.6%

Neolithic-Chalcolithic Iran

"Based on Neolithic and chalcolithic period samples recovered from Northwest Iran. The farmers from the Zagros mountain Iran region descended from one of multiple, genetically differentiated hunter-gatherer populations in southwestern Asia.  They are estimated to have separated from Early Neolithic farmers in Anatolia some 46,000 to 77,000 years ago, and show affinities to modern-day Kurd, Iranian, Pakistani and Afghan populations.  The Neolithic Iranian references used for this component, were recovered from the Kurdistan region of Iran, and appear to be around 9000 years old. The Chalcolithic Iranian references have been dated to around 5000 years old.".

My Neolithic-Chalcolithic Iran admix is 7.6%

Neolithic-Bronze Age Levant

"Based on neolithic and bronze-age period samples recovered from the Levant area in the Middle-East. The references for the bronze age Levant farmer (BA) samples were recovered from the Ain Ghazal, Jordan area and were dated to about 4300 years ago.  The first farmers of the southern Levant (Israel and Jordan) and Zagros Mountains (Iran) were strongly genetically differentiated, and each descended from local hunter-gatherers. By the time of the Bronze Age, these two populations and Anatolian-related farmers had mixed with each other and with the hunter- gatherers of Europe to drastically reduce genetic differentiation. The impact of the Near Eastern farmers extended beyond the Near East: farmers related to those of Anatolia spread westward into Europe; farmers related to those of the Levant spread southward into East Africa; farmers related to those from Iran spread northward into the Eurasian steppe; and people related to both the early farmers of Iran and to the pastoralists of he Eurasian steppe spread eastward into South Asia.".

My Neolithic-Bronze Age Levant admix is 4.4%

Eastern Non-African

"Eastern Non Africans (ENAs) are one of the earliest splits from humans that migrated out of Africa to the Near East around 100,000 years ago. It is believed that ENAs split from the population in the Near East around 50,000 years ago. Populations such as the Andamanese Onge and Papuans are modern descendants of ENAs. The ENA component here is based on Papuan references.".

My Eastern Non-African admix is 1%

Discussion

1. Timeline.  for what period in the past is this ancestry being weighed?
2. Population.  How are the proposed admixture populations quantified and distinguished?  Where did they live?  Are they associated with any archaeological culture?  What are the references - are they based on ancient DNA or inferred in modern populations?  What previous populations were they admixed from?

Admixture is repetitive.  What we are looking at are ancient migrations, population expansions (sometimes marked with culture, sometimes not), admixture - sometimes with strong sex bias, displacements - all in prehistory, long before any written record.  It's like rows of jars of mixed sweets, resulting by mixing, and remixing.  Even siblings can have slightly different "flavours".

How does the above K11 compare with some of my previous ancient admixture calculations?  How about the K7 Basal-rich for example?  It looks at a different admix, from an earlier time, but do they make sense together?

David Wesolowski's K7 Basal-rich test

Villabruna-related

The Villabruna cluster represents the DNA found in 13 individuals in Europe from after 14,000 years ago.  They were Late Ice Age hunter-gatherers.  They appear to have links with the Near East.  The current thought is that they replaced earlier groups of hunter-gatherers in Europe.  The DNA of people in the Middle East and Europe pulled together at this time, and they may represent an expansion from the South-East.  Much of the Aegean Sea would have been dry, with low sea levels (glaciation), so the migration may have been easy.  It is believed that they had dark skin, and blue eyes.  They were possibly, the last hunter-gatherers of Europe and the Middle East.  They may have contributed to our DNA both through or either, later Asian or European admixtures.

David gives the English average as 56.7%.  My result is 57.1%

Basal-rich

The Basal Eurasians are a hypothetical "ghost" population derived from DNA studies.  It is suggested that they splintered from other modern humans 45,000 years ago, presumably outside of Africa, somewhere around the Middle East.  They significantly contributed DNA to the Early Neolithic Farmers of the Fertile Crescent and Anatolia, and consequently, on to all of us modern West Eurasians.  

 David gives the English average as 26.5%.  My result is 28.8%

Ancient North Eurasian

Another Ice Age hunter-gatherer "Ghost" population, but this one has been associated with human remains and an Upper Palaeolithic culture (Mal'ta-Buret') at Lake Baikal, Siberia.  We know that it significantly contributes to modern West Eurasians, through earlier admixture on the Eurasian Steppes.  Copper Age pastoralists then carried it westwards into Europe with their later expansion.

David gives the English average as 16.6%.  My result is 14.0%

Living DNA produced their first update.  An update by a "DNA for Ancestry" business can sound like an admission of failure.  To some, it could sound like a recall due to product failure.  "Your previous ancestry was a mistake".  This only applies if you have bought into some marketing campaigns, that autosomal DNA tests for ancestry actually work even close to 100%.  Surprise, they don't!  They are cutting edge, in development, and far from accurate below a Continental level. They are still somewhere in the twilight between being nothing more than a genetic lottery, and actually becoming a tool that is useful.  Therefore "updates" are to be welcomed.  They are a sign that the business wants to improve the test accuracy.  That is to the credit of Living DNA.

My latest results?  First of all, a quick recap on my actual ancestry, as supported by family history, local history, ethnicity, and by a traditionally researched record based family tree that includes over 270 direct ancestors over the past 380 years.  I'm English.  Indeed, all of my direct ancestors, appear to have been South East English.  More precise, I'm East Anglian.  On family history and recorded genealogy, I'd suggest that between 75% and 85% of my direct ancestors over the past three centuries were East Anglian, almost all from the County of Norfolk.  Others on my father's side, if not in East Anglia, still in Southern England.

That I feel, makes me an interesting subject for ancestral auDNA testing.  You see, my ancestry is very localised here in South East England.  DNA tests such as 23andMe that claim to accurately plot ancestry over the past 300 - 500 years should get me.  But they don't.  This is because their algorythms, and reference data set designs fail over different ages.  They also (although they sometimes deny it), fail to discriminate against older population background.  We East Anglians and South East English have been heavily admixed with non-British populations on the European Continent.  Not so much over the past 500 years, so much as over the past few thousand years.

The new Results.

Below are my Living DNA regional ancestry, based on Standard Mode.

Below are my Standard Mode results broken down into sub regions.

Below is a table, comparing my recorded ancestry, with my early Living DNA results in Standard, now my revised results.

Living DNA has now introduced two new modes of confidence called complete and cautious modes.  First the Complete results:

Below are my Complete Mode results in regional:

Below are my Complete Mode results for sub-regional:

Now the Cautious results:

Below are my Cautious Mode results in regional:

Finally, below are my Cautious Mode results for sub-regional:

Conclusions

No auDNA test, by any DNA-for-ancestry company has yet come close to assigning me 100% English or even British.  They don't get me.  23andMe gives me 32-37% "British & Irish".  FT-DNA gives me "36% British".  Therefore, to be fair, Living DNA, giving me 70% "Great Britain or Ireland", give me the best result.  However, Living DNA has started out with the largest, best quality British data-set of any DNA-for-ancestry company, and is often accused of a bias towards Britain in it's results.  If so, then my 70% still looks weak.  They are planning on producing similar quality data sets soon for Ireland, Germany, then France. Therefore any results, will as I started out saying at the beginning of this post, be perpetually progressive.  Businesses that do not improve data sets or algorithms, will not get any better.  They are not progressive.

I get Southern European in other tests besides this one.  Living DNA points to Tuscany.  FT-DNA before a recent update gave me 32% Southern European, although they have revised this down to a little noise from South-East Europe!  23andMe gives me 2% Southern European - but this appears nothing unusual for an English tester.  None-the-less, I am interested in trying to better understand, why some of these tests give me this "Southern European" admixture, for which my family history, local history, and recorded genealogy has absolutely no account.  It equally reflects in ancient calculators that give me a little bit more Neolithic Farmer than for other English, which on average, already have a little more Neolithic Farmer than other British or Irish populations do.

The New Complete and Cautious Modes

How do I feel about these?  At Sub-Regional level, the Complete mode starts to get silly.  For the first time, Living DNA at this level, starts to even suggest some ancestry from Wales, SW Scotland, and Northern Ireland.  Only small percentages - but I just don't buy them.

However, the Cautious Mode, I start to like.  My British ancestry doesn't increase, but it looks more realistic, although with strange enigmatic suggestions still of Italian ancestry in the mix.  At Sub-Regional level, Cautious Mode also looks a little more likely.  My East Anglian remains at 37%, I however, lose Lincolnshire (which does exist in my record), but retain Cornwall.  I think Cornwall unlikely - however, there is just a small hint that something could be there, in surname evidence of a brick-walled great great great grandparent.  So maybe, just maybe.

East Anglia

I seriously doubt that my East Anglian ancestry over the past 300 years genuinely falls much below 75%.  Living DNA only appears to recognise a half of it at 37% - but they claim to be easily able to identify East Anglian DNA.  They call it "Distinct" because of it's high levels of Continental admixture.  They have admitted that based on their early data sets, that it was hard to separate from Germanic.  I don't know why it isn't stronger in my results.  I honestly do believe that the test underplays it on my results, even though it is the strongest of any population in my test results.  My East Anglian ancestors lived mainly in Eastern, Central, and Southern Norfolk.

Living DNA also provide a chart of the Continental "contributing regions" to East Anglian ancestry:

Finally, a chart breaking down their proposal of my British ancestry at Cautious mode:

I'm not disappointed with Living DNA.  That it does identify me as 37% East Anglian is I believe, incredibly good, and far advanced over any other DNA-for-ancestry test.  I'm looking forward to more updates in the future.  Well done Living DNA.

After a four month wait, my initial results have arrived today from Living DNA.  The wait has, I feel been understandable for a launch company.  The results are still limited to standard mode only.

Living DNA Standard mode

My initial response?  Enthralled and highly impressed.  A little disappointed that the East Anglia percentage was not higher.  I suggest 77% based on my documentary record.  Living DNA gave it 39%.  I still find that a very good result.

However... let's get this into perspective to 23andMe and FT-DNA tests.  Documentary evidence suggests that I am 100% British over the past 300 years.  23andMe said 32%.  FT-DNA said 36%.  Living DNA gets it so much closer at 74%!  That is a whole lot more accurate.

What about the remaining 26% on regional level, where do Living DNA say that comes from? All European.  It suggests 9.7% unassigned European, 9.6% Tuscan (Southern European), 4.6% Scandinavian, and 2% "Germanic".  The Tuscan is interesting, but I'm not convinced yet that it is not ancient and population based.  The Scandinavian is also most likely ancient - in my opinion.

Two things please and impress me about my results on the sub regional level:

1) Based on documentary research, I estimate that 250 years ago, 77% of my ancestors were in East Anglia.  Living DNA indeed, sees it as by far my largest sub regional percentage.  At 39%, a little low, but very impressive.  They correctly identified me as East Anglian.

My next main region, in my Family Tree, I have circa 12% ancestors from "South Central England".  Living DNA saw this, and it is indeed, my second  largest percentage at sub regional level.  I get South Central England with 7.5% - incredible.  The small "South England" would also tied to this line.

Then I get 5.4% South-East England.  It could be over spill from the East Anglia ancestry, but I do have one 3xgreat grandfather Shawers In London, that I do not know the origins of.  I wonder now?

Then it's "Lincolnshire" with 4.8%.  Brilliant!  I had a 3xgreat grandfather from the southern parts of Living DNA's Lincolnshire sub region.  That fills my documentary record almost perfect.  The small "Central England" percentage would also tie to this line.

Then follows a number of low percentages from all over Southern and Eastern England.  They might tell a story, or might not.  Surprisingly Cornwall and Devon show up in low percentages, as does Yorkshire.  Did my Shawers line actually come from one of those regions?  I have seen Shawers in Devon, Cornwall, Shores in Yorkshire, and Shawers in Lancashire.

2) What is excluded can also demonstrate the accuracy of such a test.  No Welsh, Northumbrian, French, Normand, Irish, Scottish, or Iberian ancestry suggested.  Not that I'd have any objection against descent from any of these, or anywhere - but that this test successfully sees that I am NOT descended from these close regions, is to my mind, a great success, and a vast improvement on any past autosomal DNA tests for ancestry by other businesses.  The truth is, that the English are so like these other populations!

On mtDNA they get my haplogroup down to H6a1a.

They have not yet completed my Y-DNA analysis.  I guess L in an English tester might have thrown them a bit.

No other DNA test has ever existed quite like this.  My initial response is - an amazing test.  The future of autosomal testing for Ancestry.

Here I map the ancestral events as recorded on my Gramps genealogical database.  These events can be baptisms, marriages, census records, etc.  The larger the dot, the more events for that particular parish.  I have modified images of Southern England from OpenStreetMap.org Copyright attribution-sharealike 2.0 generic.

My Mother's Ancestral Events.

This includes the recorded events for my mother's 134 recorded direct ancestors and siblings.  As you can see, her known ancestry over the past 330 years has been incredibly localised!  All English.  All East Anglian.  Almost entirely in Norfolk - with one line drifting back to nearby Suffolk.  An incredibly dense cluster in East Norfolk, around the River Yare in Broadland.  Sure enough second cousin and third cousin marriages have been detected in her tree.

My Father's Ancestral Events

This includes the recorded events for my late father's 116 recorded direct ancestors and siblings.  A little more travelled over the past 330 years, although I feel that the events record has a bias in research to show this - as indeed, I estimate his known Norfolk ancestry over the past 330 years to amount to at least 70% of his combined heritage.  Nonetheless, some of his lines trace back temporarily to London, then back mainly to Oxfordshire and the Thames Valley.  All South-East English again.

None of this makes my family any more special than any other family anywhere else in the World, with any type of recent heritage and admixture.  Indeed, the English are a particularly admixed population. However, in testing commercial DNA tests for ancestry, I feel that we offer a good reference sample of SE English, and even East Anglian Norfolk.

I'm particularly interested in how these commercial DNA companies are failing to discriminate ancient or population admixture, from recent (350 years) family admixture.  Some populations they are able to detect with some certainty and accuracy.  However, others such as the English, not at all.  They are unable - despite their claims otherwise, to break recent autosomal admixture on lines over the past ten generations, from earlier, sometimes much earlier population admixtures.

I'm looking forward to seeing if the new Living DNA test fares any better, with it's rich British data set.

Well, the kit arrived in good time!  Living DNA only just announced yesterday afternoon, on their Facebook Page, that the first batch of kits were ready for despatch - and it was here in Norfolk this morning!

Unfortunately, I can't activate it until the service goes live on Friday midday.  So I've got to wait a few days.  Ugh.  I also notice that the return address for the completed sample, is back to their base in Frome, Somerset.  So I'm guessing that they'll pile them up until they ship them to the lab in Denmark.  Must be patient!