DNAnigma: Autosomalgeddon; or, “Them’s that got shall have. Them’s that not shall lose.”

Day 2. Ancestry DNA has rolled out its better “mousetrap.” A new and “improved” way of identifying genetic relationships. I dropped from 80 pages of matches to 17. I lost a known 4th cousin (whom I match at 23andme, FTDNA and Gedmatch). I’ve lost nearly all the distant Euro-descended matches that lent credence to speculation about some of my white ancestors. I’ve lost all but three “shaky leaf” shared ancestor hints. I have no Circles. I’ve gained some new matches. Most have private trees or no tree at all. None have shaky leaves. None share my surnames.

This is not a win.

I knew the new analysis would disproportionately negatively impact non-whites, adoptees or those who otherwise have limited information about their ancestry, and I’m waiting vainly for an authoritative acknowledgment of that fact. All I’m seeing are cheery reassurances that this really is for the greater good, you’ll see. These comments seem blind to the realness of the loss of so-called “false negatives.”  This is privilege on display. For people whose genealogies descend in orderly, documented ranks, free from slavemaster paternity or undocumented marginalized others — onward and upward.

For the rest of us?  This is not a win.


DNA, Maternal Kin, Paternal Kin

DNA Definites: Ancestry (under)estimates.

Of my zillions of matches at Ancestry DNA, to date I’ve able to document 13 of them. Four were cousins I already knew; one was a cousin I conjectured, but couldn’t prove; and two others are from family lines I knew, though I did not know the match. I am related to the remaining six — the most distant matches — via late colonial or early antebellum-era white ancestors previously identified but unproven.

The chart below shows Ancestry DNA’s estimates of my kinship to these 13, as well as our actual relationship. Ancestry tends to underestimate relationship slightly in matches closer than five degrees, and I try to keep this in mind when speculating about my mystery matches.

Match Ancestry Estimate Actual Relationship
W.H. 3rd-4th cousin 2nd cousin, once removed
G.J. 4th-6th cousin 2nd cousin, once removed
H.B. 4th-6th cousin 3rd cousin, once removed
S.D. 4th-6th cousin 3rd cousin
G.P. 5th-8th cousin 3rd cousin, 3x removed
E.G. 5th-8th cousin 4th cousin
B.J. 5th-8th cousin 4th cousin
G.L. 5th-8th cousin 5th cousin, once removed
J.W. 5th-8th cousin 5th cousin, once removed
D.M. 5th-8th cousin 5th cousin, once removed
J.B. 5th-8th cousin 5th cousin, twice removed
E.D 5th-8th cousin 6th cousin, twice removed
L.B. 5th-8th cousin 7th cousin
DNA, North Carolina, Paternal Kin

DNA Definites, no. 15: Henderson.

I spotted the match on Ancestry DNA back in February. A German surname. A family tree largely filled with what appeared to be Germans and Frenchmen. But interspersed among the list of ethnic origins — Nigeria, Mali, Ivory Coast/Mali, Senegal …? I examined the tree a little more closely, and — there — could it be? A name I recognized. A rather common name, but one that matched that of a grandson of Ann Elizabeth Henderson Simmons, my great-great-grandmother Loudie Henderson‘s sister.

I reached out.  I also emailed a cousin, the daughter of the named man’s sister, if she thought her uncle had offspring of that age, in that place. It was very possible, she said. This uncle had not been in touch much. He was believed to have assumed a sort of liminal identity — not quite white maybe, but far from black. He had married several times, she thought, and had died in California.

Months passed.

Then, at the beginning of September, I heard back from E.G. He doesn’t know much about his grandfather, the match, but wondered if I did. Yes, I replied, I do. Ancestry (with its usual underestimating) pegged E.G. and I as 5th to 8th cousins, but we are 4th. Our great-great-grandmothers were sisters. He is the first Henderson relative I have matched beyond my double-cousins (whom I match more closely as Aldridges), and the first in my own Lewis line beyond my immediate relatives.

DNA, Maternal Kin, North Carolina

DNA Definites, no. 13: Nicholson.

For all I malign the inadequacies of Ancestry DNA, it has yielded up distant-relative matches in a way that 23andme has yet to touch. I did a search for “Nicholson” among my matches, and J.W.B. popped up. As I scanned his family tree, the name that snagged my eye first was “Jehu Idol.” I knew that name. I LOVED that name. So economical. So Biblical. So 18th century. And the husband of Hannah Nicholson, sister of John Stockton Nicholson Jr. and half-sister of James Nicholson. J.W.B. and I are 5th cousins, twice removed.

DNA, Maternal Kin, North Carolina

DNA Definites, no. 12: Van Pool.

Iredell County, North Carolina, was settled in the mid-eighteenth century largely by Scots-Irish and German immigrants pushing down the spine of the Shenandoah Valley from the mid-Atlantic colonies. Interspersed among them, of course, were the omnipresent English, but also here and there a Dutch family. Such were the Van Pools, who settled first in Maryland before heading south in the mid-18th century.

Ancestry DNA estimates D.F.M. as my 5th-8th cousin. I would not have noticed her name among my hundreds of distant matches, but Shared Ancestor Hint — a comparison of family trees, and Ancestry’s best feature — brought her into sharp focus. To be precise, we are fifth cousins, once removed, both descendants of John Van Pool and his wife, Elizabeth Peyster Van Pool. I am descended from daughter Nancy Van Pool, who married Samuel McNeely, and D.F.M. from son David Van Pool. Nancy and Samuel McNeely’s son John W. McNeely fathered Henry W. McNeely, my maternal grandmother’s maternal grandfather.

Ancestry also estimates L.B. as my 5th-8th cousin, and I found him, too, via Shared Ancestor Hint. We are in fact 7th cousins, descended an equal number of generations from Jacob Van Pool and Elizabeth Hampton via sons John Van Pool and Henry Van Pool.


DNAnigma, no. 14: small worlds.

Last night, idly perusing my most recent Ancestry.com matches, I ran across a screen name I recognized. I zapped an excited Facebook message, and this morning here’s what M.S. posted:

“Cost of being an Ancestry.com member: $20 a month. Cost of getting your DNA analysis: $100. Finding out that Lisa Henderson, whom you already admire and love like a sister is actually related to you when you were just asking her to help you sort through all the genetic genealogy: PRICELESS!  Wow!”

I’ve known M.S. and his sister A.S. for nearly 30 years — A. and I went to college together — and our friendship has deepened in the last several years. (A.S. has an amazing Tumblr, Hey to Your Mama N’em.) A few years ago, I did some rudimentary genealogical research for them and have encouraged M. and offered advice as his own research has deepened and as he began his foray into DNA testing. The last thing I ever imagined was that we’d share chromosomes! I have no idea what our link is, but neither of us will leave a stone unturned trying to find out!

DNA, North Carolina, Paternal Kin

DNAnigma, no. 13: high-school classmates?

Over the weekend, I did one of my infrequent checks for matches at Ancestry DNA. I found a new match to C.B., an estimated 5th-8th cousin. Heaving a sigh, I idly checked his family tree — and immediately recognized many of his surnames as common to Wilson County, my birthplace. I looked a little more closely at his profile, and … I’ll be damned. His daughter was my high school classmate! How in the world are we connected?

M.W. is the second Beddingfield High School grad that I’ve matched in Ancestry or 23andme. The other was a classmate of my sister. I have no clue how we match M.R. either.

I can assume the C.B. match is on my father’s side, as is M.R. I also assume that it is through an Anglo ancestor. What throws me is that I don’t know of any white ancestors from Wilson County or northern Wayne or southern Edgecombe Counties, from which Wilson was created. Clearly, I have one, or some, though, as these and a couple of other Wilson County matches attest. The most likely conduit is through my Artis-Seaberry-Hagans, who had obvious Euro ancestry about which I know nothing and who lived in northern Wayne County.

An initial exchange of messages with M.R. has fallen silent, but I’m hoping a collabo with M.W. will get me somewhere.