P1 and P2 are not always the same!

So Ancestry launched a new DNA testing feature this week that caused quite a buzz.

Called SideView, it gives testers a view of what Ancestry calls “ethnicity inheritance” — the estimated portions of each geopolitical region included in its admixture percentages that were inherited from each side of the family, whether paternal or maternal, without having tested either parent.¹

SideView presents that data both in a graph and in a chart, identifying the parents of each tester as P1 and P2.

What SideView doesn’t do — what it can’t do — is tell us which parent is which. Ancestry is very clear about this: “Even with the help of matches, your DNA data doesn’t tell us which parent each half of your DNA came from. We can split up your DNA according to the two parents who passed it down, but we still can’t connect each half to a specific parent.”²

Which, of course, didn’t stop people from making assumptions.

In particular, the assumption that P1 was always paternal, and P2 was always maternal.

Or — sigh — vice versa.

Based on their own, singular, one-of-a-kind results.

Um…

Nope.

Indulging in assumptions without requiring supporting evidence is one of the ways we as genealogists get ourselves into trouble — we assume that a man and woman who share a surname are husband and wife when they could be cousins or siblings or totally unrelated, or that the widow named in the will was the mother of all of the deceased man’s children, when she could be the mother of some or all or none of those kids.

Assumptions when it comes to DNA testing are no different. We get ourselves into a jam if we assume that two people who share — say — 330cM of autosomal DNA are half-first cousins when, in fact, the odds are almost exactly the same that they’re half-first cousins once removed.³

Our field’s best practices specifically warn us against assumptions, and remind us of the need to seek out supporting evidence before simply accepting them. And where we can’t find supporting evidence, we don’t incorporate assumptions into our genealogical conclusions.⁴

So… let’s look at the evidence. My older sister and I have both DNA tested. At Ancestry, we share 2818cM of autosomal DNA — that’s 100% odds of full sibling, according to the Shared cM Project tool⁵ — and at Family Tree DNA we can see that we share an entire X chromosome, as paternal sisters must, and the same maternal haplogroup, as maternal sisters must. The DNA is unequivocal that we are full biological siblings: there’s no chance that we’re half-siblings or any more distant relations. In short, we share both parents.

Our parents have fairly distinct geopolitical ancestry: our mother’s people were pretty much from the British Isles, with — we think — a little French tossed in way back; our father was born in Germany and his ancestry is solidly from the areas that today are Germany and north Germany as far back as we can find records. Only one geopolitical area in our DNA makeup has much of a potential overlap: some of our mother’s England and Northwest Europe may be that French, and some of our father’s German heritage may show up as Northwest Europe and get lumped into that same area.

So we’ll look only at the geopolitical areas that are really distinct: our father should have all the Germanic Europe and all the Sweden and Denmark; our mother should have all the Scottish, Irish and Welsh. And in fact Ancestry does put all the Germanic Europe and all the Sweden and Denmark on one side and all the Scottish, Irish and Welsh on the other.

Here’s the German / Sweden / Denmark assignment for the two of us — and this has got to be our father’s side:

And here’s the Scottish / Irish / Welsh assignment for the two of us — and this has got to be our mother’s side:

Notice that P1 and P2 are not the same for both of us.

Even though we have the same father and the same mother, P1 for me is paternal and P2 maternal, and it’s the exact reverse for my sister.

And you don’t have to take this one example as proof either. Blaine Bettinger ran a poll in his Genetic Genealogy Tips & Tricks group on Facebook, asking those who could identify which parent was which to indicate which parent was shown as P1 and which as P2. As of today, there are 1,177 responses to the poll. Of those, 617 — or 52.4% — report that P1 for them was paternal and P2 was maternal, while 560 — or 47.6% — report that P1 was maternal and P2 was paternal.

The moral of the story of course is simple: no assumptions.

P1 and P2 in SideView are assigned without regard to which one is paternal and which one is maternal.

Really.

Cite/link to this post: Judy G. Russell, “No assumptions!,” The Legal Genealogist (https://www.legalgenealogist.com/blog : posted 17 Apr 2022).

SOURCES