Friday, July 7, 2017

DNA Research

I’ve been fortunate that my family has a well-documented history, and I’ve been able to research most lines of my family tree back many generations.  Because of that I was never too interested in having my DNA tested, as I already knew where my ancestors were from.  Additionally, I was skeptical about how accurate DNA tests could be in determining exact ethnic origins. And my interest in genealogy is mostly a focus on the stories of peoples' lives - so I don't know if I even care where my ancestors happened to live hundreds or thousands of years ago. 
Siblings Percy and Violet Roberts, circa 1897. They were the first cousins
of my great-grandfather and all of them had the same amount of Asian (or
Native American) ancestry that they inherited from their mothers
(who were twins).

This photo shows the genetic variety that can appear in one family:
one sibling looks more Asian and one looks more European
 
Recently though, I decided to take the plunge and have my DNA tested.  Mostly I wanted to participate because I was curious about the process and I have an interest in the science of genetics.  Additionally, I wanted to test the accuracy of my own research.  Does my research and evidence match my DNA???

As a result of 15+ years of research, I have been able to trace most lines of my family tree back to their origins in the Old World. I knew that I was mostly (perhaps 90%) English, German and Scottish. I also knew that the rest of my ancestry (the last 10%) was  Danish, Dutch, Welsh, French, Swedish, Irish, and Swiss.  Aside from those northwestern European ancestors, I also knew about (unproven) legends of trace amounts of Native American and even Moroccan ancestry.  By and large though, most of my ancestors would have been considered Anglo-Saxon Protestants.
 

I had my DNA tested through AncestryDNA, which provides autosomal DNA testing services.  When I got my results back, I was actually quite surprised.  The main thing we both agree on is that I am mostly from Great Britain (including England and Scotland).  Yet, it was surprising that my second place hit was Scandinavia.  Other than one great-great-great-grandfather who was Danish (1/32nd or 3.125%), and a couple of very distant ancestors from Sweden (2/256th or 0.8%), I had no other (known) Scandinavian ancestors.  It was also odd that “Europe West” (which includes Germany) shows up low at only 10%, as my documented German ancestors were fairly prolific (at least 23/128th or 18%).  Additionally, the Native American and Moroccan theories were apparently proven false; as I showed up at 0% in both regions.

summary of my Genetic Ancestry results from AncestryDNA
What was most surprising to me though was that the results indicated that I am 9% West Asian, or more specifically from the Caucasus (which is an area above the Middle East that includes Armenia, Georgia and Azerbaijan; and neighboring countries like Turkey and Iran).   It’s not especially weird to have non-European ancestors, but in this case – it does seem a little odd considering how much I know about my ancestors.  My family has fairly deep roots in America, as most lines of my ancestors settled here in the 1600s or 1700s.  I’ve been able to trace all branches of my family tree back fairly far.  In fact, my most recent “dead end” ancestor was born in about 1800, in Kentucky.  I can safely say that in 1800, all of my ancestors were living in America, Germany, Denmark or Wales. 
  
Because of the historical unlikelihood of an Anglo-Saxon person interbreeding with a West Asian person in colonial America and because of my relatively high amount of that DNA, it’s likely then that my Asian blood came from one branch of my family tree, and from a relatively recent ancestor.  Assuming that we inherit equal amounts of DNA from each of our ancestors in each generation (which we don’t), it’s most likely that the DNA would have come from a single great-great-grandparent (1/16th or 6.25 percent of that generation of my ancestry). Although genetics don't really work that way, it gave me a clue for where to start looking. 

Luckily, I know quite a bit about my ancestors and have photographs of many of them, including all 8 of my great-grandparents, all 16 of my great-great-grandparents and quite a few of my great-great-great-grandparents.  My suspicion for this bloodline quickly fell on my great-great-grandmother Josephine Martin.  Compared to other branches of my tree, I know comparatively little about her ancestry.  In addition, I have several photos of her – and to me – she had facial features and coloring that don’t quite look Western European.  And there is a family story among descendants of her sister that they were Native American (perhaps as an explanation for their appearance).  In spite of that, she and all of her family were always classified as white in contemporary records.  All of her family had English or German surnames and were admittedly racist.  (Josephine was a fascinating person in her own right and something of a bad-ass.  I wrote an earlier article about her life: pioneer feminist.)

So what is going on here?  And how would I prove or disprove it? 

My great-great-grandmother Josephine Martin, circa 1880s,
who was apparently the carrier of a significant amount
of Asian DNA.

Percy and Violet Roberts (in the photo
above), were her niece and nephew
Luckily, AncestryDNA has a feature that compares your DNA results to other Ancestry members who have had their DNA tested to see if and how closely you are genetically related (based on the amount of centimorgans you share across DNA segments).  In addition, you can then compare DNA matches' results to your documented family trees to potentially aid in proving or disproving biological relationships.  It turns out I have quite a few cousins from most branches of my tree that have had their DNA tested.  Browsing through their results was interesting and also seemed to prove what I suspected; Caucasus DNA is unusual among white Anglo-Americans.  Then I finally found other relatives with Caucasus DNA; five different cousins who are descendants of Josephine or her twin sister.

This all seemed fairly conclusive: my West Asian DNA came from my great-great-grandmother Josephine Martin.  Due to the amount of West Asian DNA inherited by myself and the five other cousins; Josephine was significantly (perhaps completely) West Asian.  However, Josephine’s life was well documented; she was born in Missouri in 1845 and her parents were white slave owners who were born in Kentucky and Virginia respectively.  And she was born during a time when emigration from West Asia to the US was incredibly rare.  How would I reconcile the DNA evidence with the other evidence?  Was she adopted?  Was a parent or a grandparent from that region?  And what in the world were people from the Middle East doing in Missouri in the 1840s?  As is usual with genealogy, research usually seems to uncover more questions than answers.  


The Caucasus


The Caucasus region is home to the Caucasus mountain range and is between the Caspian Sea and the Black Sea.  Historically, the region was culturally and ethnically diverse.  For much of its history, the region was under the control of various empires, including the Persian Empire [Iran], the Ottoman Empire [Turkey] or the Russian Empire.   Today, the Caucasus region is made up of three countries: Georgia, Armenia and Azerbaijan and it borders Russia, Turkey and Iran.  The region was important throughout history; it was at the border between Europe and Asia and was at the crossroads between the Christian and the Muslim world.  Because of its strategic location, it was frequently the site of wars and persecution of religious minorities.   Perhaps it was because of this tumultuous environment that my unknown ancestors left their homeland.  

Map of my DNA results, showing the Caucasus
area in blue.
Curiously though, my Caucasian ancestors (or their descendants) must have arrived in the US sometime before 1845.  But widespread emigration from that part of the world really did not happen until the late 1800s or early 1900s.  Based on my limited knowledge of that region’s history though, the most likely country to have had pre-1845 emigration would have been Armenia.  For example, there was a small wave of Armenian immigration to America during the 1830s as a result of oppression of Armenian Christian minorities in the Ottoman Empire, and there is documentation of individual Armenians in America back to the early 1600s.  Another tempting clue related to this theory is that Josephine had a cousin (on her mother’s side) that was named Armenia Crobarger. Was she named after her ancestral place of origin? It’s compelling but unfortunately, Armenia is just a guess and we will probably never know for sure where exactly my Caucasian ancestors came from or really anything about them.

It is of course possible that my ancestors did not directly come from the Caucasus to the US.  It is historically more likely that they would have first moved to closer regions, like Russia, Eastern Europe or Southern Europe before indirectly making their way to America.
  

Josephine’s Documented Family


According to various surviving records, Josephine was born on June 3, 1845 in Platte City, Missouri; near what is now Kansas City.  At the time, this was at the edge of the frontier of America; they were literally just a few miles away from unorganized frontier territory and the start of the Oregon Trail.  

Her father, William Martin, was born in 1815 in Kentucky.  His parents, Zadock Martin and Susannah Brown, were born in North Carolina and Virginia respectively.  They moved to Missouri when William was a young child, and it was there that he grew to adulthood.  The paternal line of the Martin family has a documented descent from an English family that settled in Maryland in the mid-1600s.  

Josephine’s mother was Harriet Crobarger.  Harriet was born in 1816 or 1817 in Virginia or Tennessee.  I do not know who her father was. Her mother though was the widow Catherine Crobarger, who raised Harriet and her siblings.  They lived for many years in Tennessee, and lived briefly in Indiana before settling in Missouri in the late 1830s; shortly before Harriet married William Martin.  The Crobarger family likely descends from a German family that settled in Pennsylvania in the mid-1700s.
  
Josephine was a twin; born alongside her sister Frances “Fanny”.  They had 5 other siblings: Catherine, Hardin, Joseph, George and Emma.  

Possible Explanations for the West Asian DNA


The DNA evidence suggests that Josephine carried a significant amount of West Asian DNA.  It is even possible that she was 100% West Asian.  However, given the fact that at least some of her documented ancestors were English or German, it is necessary to hypothesize what might have happened.  

Below are some of the theories I have come up with:

1. Josephine and her twin sister Frances were adopted by William and Harriet Martin.  
2. Josephine and her twin sister Frances were a result of infidelity; children of Harriet Martin by an unknown West Asian man who was not her husband.  
3. Either William Martin or Harriet Crobarger were adopted or born to their mothers as a result of infidelity
4. One or more of Josephine’s four grandparents were ethnically West Asian:

4a. Her paternal grandfather, Zadock Martin.  He was born in 1789 in North Carolina.  Although I don’t technically have proof of his parentage, he was most likely the son of another Zadock Martin, and descended directly from an English family that lived for about 100 years in Maryland.  Additionally, I also have DNA matches with two distant cousins who are documented as descendants of his brothers – and neither of them show up with Caucasian DNA, although that doesn’t necessarily mean they didn’t actually have Caucasian ancestry.  (This likely is an indication that I AM biologically/genetically related to this side of the family and that the West Asian DNA did NOT come from here.) 

4b. Her paternal grandmother, Susannah Brown.  She was born sometime during the 1780s in Virginia and was raised in Kentucky.  I know nothing about her origins, except the names of two siblings and that she was possibly the daughter of Frederick Brown. 

4c. Her maternal grandfather, name unknown.  This is the easiest guess because we know close to nothing about him. All we know is his supposed surname – Crobarger – and the fact that he apparently lived in Virginia and Tennessee as an adult.  The Crobargers as a whole were German, and settled in Pennsylvania in the 1700s.  Perhaps though he was not a Crobarger at all.  Maybe it was his wife/partner – Catherine Crobarger – who was really born into that family.  All we know is that in 1830 she was living in a household with her four children, enumerated as Katharine Croborger.  For the next 45 years, she used the name Crobarger and was never explicitly described as a widow or wife of anyone.  Maybe Crobarger was her birth name, and she had a relationship with an unknown man that she did not marry, who was the father of at least one of her children.  

This theory seems even more likely when I realized that Catherine Crobarger shares a birth date (December 16, 1777) with Christina Crobarger, wife of George Pence. Christina was the documented daughter of Michael and Eva Crobarger, who I know that my Crobarger family had some kind of connection to. If these two women were twins, it would make sense because twins were definitely common among the descendants of both of them.

If this theory is true, then the unidentified father of Catherine's children was in the southwest Virginia/northeast Tennessee area between about 1800 and 1820. 

4d. Her maternal grandmother, Catherine Crobarger. Assuming that her husband really was a Crobarger, we know nothing about her maiden name or origins except that she was born in 1777, supposedly in Pennsylvania; and her first known child was born in Virginia in 1808. In one census she indicated that both her parents were foreign-born.  Although most sources list her birthplace as Pennsylvania, one source (the 1850 census) just lists her birthplace as “O.”  The census enumerator probably meant Oregon or Ohio, but neither of those is possible.  Could “O” have stood for something else entirely, like Orient or Ottoman Empire? 

Conclusion?  Unfortunately, there is no way to prove any of the above theories.  My educated guess though is that Josephine and her twin sister were in fact the biological children of William and Harriet Martin. And it is most likely that the West Asian DNA comes from their mother, Harriet Crobarger.  This also fits with the clue that the name "Armenia" was used in the Crobarger side of the family and that the Crobarger line also contained family legend of Native American ancestry.  In addition, although all of these families were southerners, Harriet's brother Francis Crobarger supported the North during the Civil War while most of the rest of the family supported the South.  It seems more likely that the family containing non-white individuals would have been against the institution of slavery.

More research is needed. 


Racism
Josephine's 10th child, David Plymale, in 1883


I am educated as a sociologist, which allows me to often look at issues through a larger lens and examine the social institutions that affect peoples' lives.  In this case, the context of Josephine’s life was relatively fascinating.  Although she was (probably) a mixed-ethnic person, she was raised by and married into white families that were admittedly quite racist. She also lived in a society where racism was not just institutionalized but also blatant and acceptable, even in the progressive North.

Her father


Josephine’s father William Martin was born in Kentucky and raised in Missouri.  He grew up around the institution of slavery and his parents owned slaves throughout his upbringing.  Years later, he supported the Democratic Party and the Confederacy during the Civil War.  Although many historians have tried to justify support for the Confederacy as part of a non-racist, political ideology (i.e. states’ rights), the reality is that those who supported slavery were conditioned to do so, and viewed black people not just as lesser people, but not as people at all.

In Missouri, the Martin family lived in the frontier of America where they often interacted with the Native Americans that lived nearby.  There is a story that has survived about how an Indian played a trick on one of William's brothers, Gill Martin, when they were teenagers.  It was perhaps because of these experiences and conditioning that in adulthood, William apparently developed a distrust and hatred of Indian people.  In fact, he served as a volunteer soldier in three different Indian wars and was directly responsible for the murder of many Indian people throughout his life.  He was a soldier in the Seminole Indian War in Florida from 1837-1838.  He was a captain in the Cayuse Indian War in Oregon from 1847-1848 and he was a colonel in the Rogue River Indian Wars in Oregon from 1855-1856. 

I am fortunate to have many surviving letters written by William during his lifetime (all of them to his friend and political ally, Joseph Lane – a general, governor and senator).  Many of the letters expressed disparaging attitudes towards other ethnic groups; including his desire to kill Indian people during war times and using the “n” word to describe Mexican people.  

William's story is even more interesting when I realize that some of his wife's relatives (who were also southerners) supported the North during the Civil War.  The sources also seem to suggest that William remained close with his various relatives; and it is interesting to see that the family remained close in spite of wars, political differences and racism, which might ordinarily tear families apart.  

Her Husband


When Josephine was 17, she left her family and moved away to become a school teacher.  The following year she married William Plymale and they started a family together.  The Plymales were probably a little less racist than the Martins.  William was born in Illinois; and the story is that his parents had moved there from Virginia because they did not approve of slavery which was legal there.  In his later years, William was a writer and considered himself a historian.  One of the papers he wrote lambasted the White man’s terrible treatment of Native Americans, while also still maintaining that Native Americans were inferior people.  His paper specifically said: “While it is true that the “white man”, and especially the Aryan or Anglo-Saxon type, possesses in the most eminent degree many of the noble, generous and admirable qualities, and is in fact the greatest and grandest type of mankind on earth…”


So although William was apparently against slavery and the poor treatment of Indians, he definitely thought that white people were superior. 

Josephine


So how does Josephine, who was apparently not an Aryan person, fit into this?  I know quite a bit about Josephine’s life, and what she was passionate about (including education, fruit farming, history, women’s suffrage, progressive politics, temperance, etc.) but absolutely nothing about her views about racism or non-white people.  Perhaps nobody – including herself – even knew she was anything but white.  Despite her appearance, she was definitely passable as a white person and that was the identity used by her throughout her life.  The surviving records suggest that Josephine was a fiery, passionate and strong person; and was not meek or subservient like women were expected to be.  While her father and husband were both Democrats, she was an outspoken Republican (which tended to be the more liberal/progressive party of the time).  She was a prominent women’s suffrage activist, worked for the Oregon State senate, and even tried to run for political office (and was ridiculed for doing so). 

It's possible that Josephine's silence on the subject and the blatant racism within her family may have partially been a result of sensitivity or resentment over their own Asian heritage. 



Photographs


I am fortunate that I have photographs of many members of this family, priceless aids in this journey of exploring potential Asian ancestry.  I have four photographs of Josephine.  In addition, I have a photograph of her father, William Martin, but none of her other ancestors or siblings.  I also have photographs of 10 of her children and 5 of her twin sister's children.

Josephine 

Below are the four photographs I have of Josephine. To me, her appearance was unusual; especially in contrast to her Anglo-Saxon counterparts. Various evidence suggest that she was partially northwest European and partially Asian or Native American.

Josephine
Josephine
Josephine
Josephine

 

 

 

 Father

Below is the only photograph I have of Josephine's father, William Martin. The image is so weird though, that it's hard to really compare it to compare it to others. That said, I really don't see any resemblance between Josephine and her father.
William J. Martin

Children

Josephine had 12 children and I have photographs of 10 of them, below. It is interesting that the majority of the children seem to take after their father; many of them had light coloring, blue eyes and his facial features. Her three youngest children were the ones that most resembled her. How does birth order play a role here?
William L. Plymale
Ada C. Plymale
Kate M. Plymale
Louis H. Plymale
Frank M. Plymale
Emaline J. Plymale
Marie F. Plymale
David H. Plymale
Victor B. Plymale
Benjamin H. Plymale
(my great-grandfather)

Niece and Nephews

Josephine's twin sister Frances also had 12 children, and I have photographs of 5 of them, below.
George W. Roberts
Joseph L. Roberts
Perle F. Roberts
Percy H. Roberts
(Percy and Perle were twins)
Violet R. Roberts



Update - A Different Test


Because I was so fascinated by these results, I decided to have my DNA tested through a different company.  I submitted my sample to 23andMe, which also provides autosomal DNA testing (similar to AncestryDNA).  My hope was that the ancestry profiles of both tests would be identical or at least mostly the same.  I was wrong.  

Comparison of my DNA ancestry composition from AncestryDNA
and 23andMe
The overall results were similar – in that they both agree that I am mostly British (57% vs 52%) and northwestern European in general (79% vs 95%), and both show trace amounts of DNA from Eastern Europe and from Finland. But they differ from there.  The AncestryDNA test shows trace amounts of DNA from Spain, Portugal, Italy and Greece, but the 23andMe test does not.


The region identified as Caucasus in the AncestryDNA test is covered by the Middle Eastern region in 23andMe. Yet the 23andMe test showed up at 0% in the Middle Eastern region.

The other discrepancy is that I showed a positive result for East Asian DNA in the 23andMe test, specifically in the Yakut region.  The Yakut are indigenous people that live in Siberia, in eastern Russia.  The reports from 23andMe suggest that the Yakut DNA entered my family relatively recently; and that specifically someone born during the 1700s was 100% Yakut.  

So what is going on here? Why does one test indicate I am partially West Asian while another indicates I am partially East Asian?  It’s tempting to suggest that one of the samples was wrong or contaminated, but both tests matched me with known relatives and the AncestryDNA test also showed the five other cousins with a similar amount of Caucasus DNA.

In reality, the different results suggest that the two different companies use different methods/processes and that the whole thing is a fairly inexact science and should be “taken with a grain of salt”.   Essentially, each company's method of predicting ethnic origin really just depends on the size and accuracy of the reference dataset that they use to compare your DNA to (and an assumption that DNA of people with the same ethnic origin will look the same).  In this case, AncestryDNA has a dataset specific to the Caucasus, whereas 23andMe does not; which seems to suggests that the AncestryDNA result is more specific and thus more accurate.  To that end, an employee of 23andMe even responded to a user's question about DNA results for Armenians and said that without a dataset for the Caucasus, the results of people from there would probably not be accurate and would instead probably show up as European, rather than Middle Eastern.  (In addition, 23andMe has a dataset specific to Yakut, whereas AncestryDNA does not; which might explain why I show results there in one test and not in the other.) 

To boil it down, your genes themselves don't identify where your ancestors came from. Instead, your DNA is just compared to other DNA samples and those comparisons are used to guess where your ancestors might have come from. 


This process overall has been educational for me, as I learned that the way we inherit our genes is random and does not necessarily match our biological ancestry.  In particular, the genes we inherit from our ancestors are not evenly distributed.  This means that in a particular generation, we probably have zero genes from some of our biological ancestors and a larger amount of genes from others. So by default, it's likely that your DNA will not match your biological family tree. 

Regardless, both test results are similar overall and they mostly agree with my own research. They also both seem to agree that I had some Asian ancestors in the relatively recent past.  And those results seem to match with photographic and oral history evidence that my family contained some non-white ancestry.  And based on the reasons highlighted in this article, those genes (whatever the ethnic origin) were probably in the ancestry of my great-great-grandmother Josephine Martin, specifically through her mother Harriet.  


A New Theory - Cherokee


While doing research on this issue, I came across theories from a distant cousin in the Crobarger line who hypothesized that our family included either Cherokee or Melungeon ancestors (based on appearance, timeline, and geographic location).  When briefly looking into this possibility, I stumbled across something that is actually quite fascinating.  Apparently a few genetic studies have found that the DNA of certain Cherokee Indians is most similar to people from the Middle East and has little in common with other Native Americans of East Asian descent.  As far as I know, all of these tests have been of mitochondrial DNA (mtDNA) of people with a documented maternal line descent from Cherokee tribal members in North Carolina. The haplogroups resulting from these tests are by-and-large common in people from the Middle East or neighboring areas like southern Europe and north Africa, and are not found in indigenous people in the Americas or East Asia.

These findings also seem to agree with the claims of some historians, that the culture and language of the Cherokee tribe was considerably different from other tribes in the southeast United States and that there is no historical evidence of the tribe before the early 1700s (even though there is documentation of other tribes much earlier than that).  As far as I know, nobody knows why or how that is the case.  Many historians though think that the Cherokees actually originated in the Great Lakes region and migrated to traditional Cherokee territories relatively recently.  That might explain the historical and cultural discrepancies, but does not explain the DNA evidence.  Various theories though have been proposed, and as far as I can tell, they fit into four main groups:

1. The Cherokee tribe does not actually have an ancient history in North America, and instead they are descendants of Jewish emigrants and other marginal ethnic groups who moved to North America in the 1600s. Over time, these settlers merged with other groups, including some local Native Americans and eventually gained a tribal identity as Cherokees. 
2. The ancestors of the Cherokee tribe moved to the New World from the Mediterranean/Middle East in a previously unknown Atlantic Ocean crossing, sometime before Christopher Columbus.  Some people even think that there is a biblical explanation, and that the Cherokees are one of the Lost Tribes of Israel.
3. The DNA testing "proving" that Cherokees have Middle Eastern ancestry is bogus, and the Cherokees' ancestors have been in the Americas for thousands of years, just like all of the other Native American tribes.
4. The existence of Middle Eastern DNA says nothing about the tribe's ancient history, and is only evidence of a high degree of admixture (interbreeding) with other groups of people. 

I am not interested in getting embroiled in this controversial issue.  But, if Cherokee DNA does sometimes show up as Middle Eastern, it would potentially solve the riddle posed by this article and would explain why some of the DNA of myself and five cousins is similar to people from the Caucasus (at least according to AncestryDNA). If one accepts this Cherokee ancestry, it makes much more sense historically than either Caucasus or Yakut.  As referenced in the article above, my non-European ancestry likely came from my great-great-great-grandmother Harriet Crobarger - which means it probably entered the family in northeast Tennessee or southwest Virginia; in the late 1700s or early 1800s.  This was exactly where Cherokee people were living and was during a time of increased interaction between Cherokees and European-American settlers.

All of this considered, my current theory is that my great-great-great-great-grandfather (unidentified father of Harriet Crobarger) was a Cherokee Indian.  He was probably born in southwest Virginia, in the second half of the 1700s, and also lived in Tennessee as an adult.  He had a relationship with a German-American woman named Catherine Crobarger, by whom he had four children.  The inference from surviving records is that they moved frequently during their relationship and were apparently not married, it probably not being legal to do so.  The Cherokee people had a matriarchal society, in which clan identity was passed through the mother.  This might also explain why their four children all used their mother's last name - Crobarger - as their own, even if the parents were married. In addition the children were all given "White" names (George, Susannah, Francis and Harriet) and they all identified as White in adulthood.  

If this theory is true, then my great-great-grandmother Josephine Martin (primary subject of this article) was one-quarter Cherokee. (And would also mean that I am 1/64th Cherokee.) And if we have identified these ancestors as Cherokee, it opens up another can of words - genetically, who were the Cherokees? 

At the end of the day, I’ll probably never know anything more about these unknown people; but the process of trying to find answers is fun and rewarding and helps me to learn more about people I am more closely related to.  And regardless of the accuracy of the tests, the whole thing is rather irrelevant anyway as humans are genetically quite similar.  And to me, the history of humanity is really a history of migration; humans have been moving around for thousands of years!


In addition to AncestryDNA and 23andMe, I also had my DNA sample run through two other prominent companies' autosomal ethnicity prediction services.  This table shows my own research compared to my estimated genetic makeup from four different companies.

Check out this recent Wire Cutter article which rated AncestryDNA as the best DNA testing service, in part because they use the largest database of samples. 

The tests described above were all autosomal DNA tests, which have their limitations.  There are a couple of other tests that are perhaps more accurate or specific, but are limited in different ways.  The Y-DNA test will test your Y-chromosome (males only) to determine your paternal haplogroup.  That is pretty accurate, but it only tests your direct paternal line (father, his father, his father, his father, etc.).  The mtDNA test will test your mitochondrial DNA - which everyone has - but is only passed down in the direct maternal line (mother, her mother, her mother, her mother etc.).  Both of these tests can provide clues to your origins, but only for a very small portion of your ancestry. In this case, neither my direct paternal or maternal lines contained the ancestry described in this article, so those tests are not relevant. 

My paternal haplogroup (Y-DNA) is I-L161.1, which is relatively rare, but is found throughout Europe.  However, it is most common in the British Isles and one theory is that the I-L161.1 haplogroup members were among the earliest settlers of Britain. They would have predated most colonizers, including the Celtics, the Romans, the Anglo-Saxons, and the Vikings, and probably would have been around for the construction of Stonehenge.  Apparently, members of this haplogroup were more likely to have survived in the peripheries of Britain - like western England, Ireland and Scotland - after being pushed out or repopulated by the centuries of conquest by various other groups.  This haplogroup identification definitely fits with my documented family tree that my oldest known paternal-line ancestor was John Wadleigh, who was born in about 1600 in southwest England.

My maternal haplogroup (mtDNA) is V9.  Like my paternal haplogroup, this one is also relatively rare, but is found throughout Europe.  Apparently it is most common in northern Europe and Scandinavia.  This fits with my documented family tree; my earliest known maternal line ancestor was Elisabeth Voss, who was born in about 1750 in northern Germany, near the border with Denmark.