Journey with mitochondrial DNA

Profile manager: Barry Smith [send private message]

Journey with mitochondrial DNA

This is the story of how I shored up a hole in my family tree and confirmed that I share the exact same copy of mitchondrial DNA that my ninth-great-grandmother Hannah Lawrence had when she was born over 350 years ago. This has been the most elating of the many projects I have completed since I became serious about my genealogy hobby, enough to prompt me to write an essay about the experience.

Mitochondrial DNA

Mitochondria are found within most human cells. They are within the ovum when a child is conceived, and when the child is an adult, the hundreds of trillions of mitochondria around their body are all copies of copies of those original structures supplied by the mother. Within each mitochondrion is a genome, thought to be a vestige of bacteria that once lived symbiotiacally within some ancient species ancestral to humans. This table gives some key differences between a human mitochondrial genome and the more familiar genome each of us carries within each cell nucleus:

What does this have to do with genealogy? Close relatives will have DNA that is detectably similar. Higher levels of similarity correspond, roughly, with closer degrees of kinship.

Consider the nuclear DNA of first cousins: they will both have inherited DNA from each of their four grandparents. Because they only share two of those grandparents, and because parents only pass down a randomly determined half of their genome to each child, it turns out that only about 1/8 of the nuclear genomes of two first cousins will be identical because they inherited those same portions of DNA from their two shared grandparents. Second cousins typically share only 1/32 of their genomes, and with a further degree of cousinship, the expected amount is again cut in 1/4. The process that causes this fast drop in the amount of shared DNA is recombination, which is part of the mechanism of copying the nuclear DNA (except the sex chromosomes, in some cases). As mentioned in the table above, recombination is not part of the process of copying mitochondrial DNA.

Since there is no recombination in mitochondrial DNA, a mother will pass an identical copy of her mitochondrial DNA down to her children. In principle, each of her children should have that exact same genome, as would each of her daughters' children and all deeper matrilineal descendants. Well, almost. If this were the case, then every human on earth would have identical mitochondrial DNA, and there would be no way to deduce degree of kinship by comparing two peoples' copies.

There are two processes that lead to changes in DNA over the generations: one is recombination, but the other is mutation. Every time a genome is copied, there is a chance that there is an error made, so the new copy will have a single letter changed, inserted, or deleted compared to the original. This is a mutation. If that copy is then used to make more copies, the mutation will continue to be copied, and it may be passed down to descendants if the mutation makes it into a reproductive cell. By comparing the mutations between different people, it is possible to estimate their degree of enatic kinship, i.e., the distance you'd have to trace back solely through their matrilines to find a common ancestor. The more distant that connection is, the more unique mutations will have appeared through those generations to distinguish their genomes.

Mitochondrial DNA testing (a.k.a. mtDNA testing) was the first type of DNA test used in the study of ancient humans and their migration. A study of mtDNA genomes from humans around the world gave signficant support to the Out-of-Africa Hypothesis — that all humans descend from a population living in Africa, part of which then spread out of Africa to populate the rest of the world. Much of the initial public interest in individual DNA testing came after the publication of the late Bryan Sykes's book The Seven Daughters of Eve, in which he invented fictional accounts of the matriarchs of the different "clans," i.e. the mitochondrial haplogroups. mtDNA has also been used in some famous studies of specific family kinship. When remains in an unmarked grave in Russia were discovered and suspected to be those of Tsar Nicholas II and his family, murdered during the revolution, mtDNA was used to prove their identities. The Tsarina and Prince Philip were both matrilineal descendants of Queen Victoria, and Philip was tested and found to have mitochondrial DNA matching some of the remains.

My Matriline

I was lucky enough to be handed a substantitive family tree on a platter. Genealogy was my maternal grandfather's favorite hobby, and is one of my father's, so all of my lines to my sixteen great-great-grandparents were already known and have proved to be readily documented. My father and grandfather had already traced my matrline back farther to my 5th-great-grandmother Elizabeth Burroughs, the daughter of Benjamin and Mary Burroughs, born around 1760 in Salem County in Southern New Jersey. Online trees have stated that mother Mary was born Mary Van Horn, but I never saw direct evidence for this assertion in the form of an original or published derivative source. Some investigation revealed that her grandson James supposedly gave a daughter the name "Mary Van Horn Burroughs," but even documenting that middle name of that much more recent woman proved elusive. After a while, I found some online trees that claimed her parents were Cornellis Van Horn and Hannah Seabrook, but without source. Mary had several descendants given the name "Cornelius," so I assumed someone had inferred her parentage from this naming pattern. Mary's profile on Geni.com also suggested that there were extant private family documents. At that point, I gave up looking at this line, being too skeptical of this evidence to want to put a lot of work into documenting Cornelius Van Horn or Hannah Seabrook.

Some years passed, and I became interested in genetic testing for genealogy. The standard tests advertised on TV just test nuclear DNA, particularly the autosomal (non-sex) chromosomes. My mother did one of these tests, and I learned how to use the data to confirm ancestors on different lines. Because of recombination, these tests have extremely limited utility for confirming ancestors farther back than about 4th-great-grandparents. It was almost certainly not going to be helpful for researching the parents of Mary Van Horn.

I then learned about mtDNA testing, and I upgraded my mother's test to include an mtDNA test in late 2017. Because mtDNA does not recombine, it can be used to probe much father back than autosomal DNA, but only on the matriline of the testee. However, mtDNA mutation rates are glacial, often with no mutations occurring for many generations in a row, and this makes them difficult to use for pushing through a total brick wall on a matriline. My brick wall, however, had a big crack — the extension through Mary's purported mother Hannah Seabrook. This was a specific hypothesis that could be tested if I could find the right living person and convince them to help out. I needed a living relative enate to Hannah Seabrook (meaning related through purely maternal connections).

My interest in the documentation of my matriline had now been renewed, and by this point I had learned of the works of John Stillwell, who wrote genealogies of many early New Jersey families. He steered me towards a little manuscript, A Brief Account of the Descendants of Capt. Cornelius Van Horne of White House, Hunterdon County, New Jersey. The author, a descendant, claimed to transcribe a record from Hannah (Seabrook) Van Horn's bible. It named a daughter Mary, born in 1733, which fit very well with Mary and Benjamin Burroughs's marriage by around 1752. So the documentary evidence was now stronger, but still very weak, and I was now convinced that it would be worthwhile to put in the effort to identify a living person to test the hypothesis.

Mary reportedly had at least one half sister. But rather than looking at her descendants, I decided I'd be ambitious and try to confirm the line back as far as possible. Digging through land records, I was able to document Hannah Seabrook's mother and grandmother. Her mother was Hannah Grover, daughter of Joseph Grover and Hannah Lawrence. The latter was the daughter of William Lawrence, an early settler at Newtown, New York in the 1650s and then an original patentee at Monmouth, New Jersey in the 1660s. Online trees claim Hannah Lawrence's mother was named Hannah Townsend (so many "Hannah"s!), but no documentation seems to exist even of her forename, and regardless, she had no daughter other than Hannah.

So Hannah Lawrence would be my most distant matrilineal ancestor whom I might be able to confirm with an mtDNA test, assuming the theory about Mary Van Horn's mother was correct. Hannah had six known daughters, and the goal was now to trace down enatic lines from each of them and hope to find a living person who would be willing to test.

Searching for a tester

I began tracing lines down from Hannah Lawrence in late 2019. At the time this is written, there are 127 profiles of females on Wikitree on enatic lines descending from Hannah. I added the majority of these myself. In a few cases, I had to fix the information commonly found in online trees based on my own research, typically in land and probate records. In many cases, I continued lines further down with private research and just did not want to spend the time to add profiles to Wikitree.

Once I had extended an enatic line down to, say, around 1900, I would find the last person on the line by doing a search of public trees on Ancestry.com. These people were recent enough that they usually weren't in that many trees, although occasionally the number of trees showing a given person would exceed 100. I would begin looking through the trees, and most often, the person would instantly be recognizable as a tangential relative — not a lineal ancestor — of the person who made the tree. When the person I searched was a lineal ancestor of the tree-maker, the people in trees born after around 1900 would tend to be shown as private. But if the sexes were identified correctly, I could still make a reasonable guess about whether the person who created the tree descendended through just women.

By early October 2020, I had a list of 15 people to reach out to on Ancestry.com who had trees that suggested they were either an enatic descendant of Hannah Lawrence or else that they might know someone who was, maybe an aunt or a first cousin. I wrote private messages to each of them, giving a brief description of how I found their tree, what my project was, and the line I proposed from the person back to Hannah Lawrence. A few people wrote back and indicated that they knew of no living, enatic descendant on the line I proposed. Most never wrote back at all.

So by November 2020, I figured that I wasn't going to hear back. I could go back to work on finding more enatic lines, but I had genealogical interests and didn't want to continue using all of my genealogy time on this project. It was put on hiatus.

Then in August 2021, one of the people I contacted, a woman in Missouri, finally wrote back (she hadn't seen the message earlier because it had been a long time since she logged into Ancestry). She indicated that she was indeed a descendant down the line I proposed and that she was willing to do an mtDNA test!

Results

I paid for the mtDNA test myself and gave the woman in Missouri control of the data and the account, only stipulating that I wanted her to let me know when the results came through since I figured if she wasn't a match, then I wouldn't be informed. (There was no need for me to try to have access to the data when it came in — if she was a match, then her data would be almost identical to my own. If she wasn't a match, then I wasn't interested in anything about the data other than the the fact that it didn't match mine.) The test went off to her in Missouri, and she was very efficient, doing the test as soon as it showed up, sending it off in late August, and then letting me know. Then on September 10th, I got an email saying I had a new mtDNA match. I get emails like this frequently — on average maybe twice a month — so it was no surprise that this was just some other person's new kit showing up, a distant match with 2 mutations separating it from my own kit. It would have been an extraordinarily fast turnaround for the lab at FTDNA.

On September 30th, I got an email saying I had a new mtDNA match. I immediately hopped over to FTDNA, and there was my cousin in Missouri at the top of the list! She is a match at a genetic distance of 0, meaning that she and my mother have completely identical mitochondrial genomes. It is ten generations down from Hannah Lawrence to my mother, and ten generations down from Hannah to the woman in Missouri, and through every single one of those connections, there has never been a mutation passed down!

It is striking how old our particular version of the mitochondrial genome is. If there had been a mutation just a few generations above my mother, then we would have very few exact matches in the world today — likely only enatic descendants of that recent ancestor. Because my particular copy came into existence hundreds of years ago, there are likely in the world today cousins with my same exact mitochondrial genome who descend from an even more distant ancestor on Hannah Lawrence's matriline, born somewhere in Europe. Indeed, one of my other matches with a completely identical genome (there are currently 15 in the system) reports that he can reliable trace his matriline back through a sequence of women who all lived in the same small region in Northhamptonshire, England back into the mid-1600s. There could be thousands of people in the world today who would show as a match to my mother if they tested, maybe even millions.

With such a large quantity of people, I suppose one could argue, "your test matched this woman in Missouri, but that doesn't say much because there are tons of people in this world who would match you — you can't rule out that she was just a match by chance, not because your theory about the lines that connect you was correct." But this logic is faulty. It is clear that the vast majority of people who test their mtDNA are not matches to my mother, even distant ones separated by 1, 2, or 3 mutations. The chances of a random person I contacted in Missouri being an exact match by coincidence are miniscule. This is evidence as strong as any name recorded on an ancient document that every single one of the 20 links in the tree separating my mother from this other woman are correct, even if the evidence is still what professional genealogists would classify as "indirect." I have now updated Mary Van Horn's connection to her mother on Wikitree as "confirmed with DNA," and I have updated my most distant matrilineal ancestor in the FTDNA database from "Mary [Van Horn?] b. New Jersey ca. 1733" to "Hannah Lawrence b. New York ca. 1660."

What next? Probably nothing, with mtDNA. It is the least useful of the genetic genealogy tools when you don't have a precise theory that you are trying to test. Combing through my match list and trying to determine how those other folks are related to me will probably be a waste of time, for now.

But maybe sometime soon, there will be reason to take up my interest in my mtDNA: the Million Mito Project is ongoing and plans to develop a much more refined haplotree of mtDNA. This will be like a big family tree, except that only mitochondrial mutations are placed on the tree, when they first develop and get passed down, so the "link" between two mutations may actually span many generations of women, rather than the single generation bridging people in a standard family tree. The haplotree has not been updated between 2016 and 2021, even though the number of tests has increased many-fold over that time. In the outdated tree from 2016, the lowest branches are mostly mutations that formed usually thousands of years ago, so the tree is useless for individual family history research. On the other hand, I have no way when examining my match list on FTDNA of knowing which of the people with one mutation different than my own are above me on the tree, and which are below. There are two possible scenarios: (1) our common ancestor is Hannah Lawrence, or one of the women below her on my matriline, and this other testee got one more mutation more recently on their line down from that common ancestor, or (2) Hannah Lawrence had received a mutation from one of her ancestors, which my mother received, and the mtDNA match is missing this mutation because they descend from a more distant ancestor of Hannah Lawrence, through an even longer line without a mutation. The Million Mito proejct may end up providing me with such information.

With such information in hand, there would be two possible projects. The less ambitious project would be to link as many descendants of Hannah Lawrence to her as possible. The more ambitious would be to try to assemble all of my close matches into connected clusters. In each cluster would be people with documented enatic connections. The number of clusters could possibly suggest how long ago our particular version of the mitochondrial genome came into existence. If it were just a few generations before Hannah Lawrence, there would not be that many clusters, and otherwise, we could surmise that the genome is very old indeed.