The good, the bad and the ugly
So yesterday was the rollout of the changes in matching systems for folks who’ve had DNA tests at AncestryDNA. And, as usual, there’s good news and bad news in the changes. Enough to warrant breaking with The Legal Genealogist‘s usual Sunday-for-DNA rule to comment today.
Let’s start by remembering what we’re dealing with here. The AncestryDNA test is an autosomal DNA test. It looks at the kind of DNA that you inherit equally from both of your parents: you get 22 autosomes (plus one gender-determinative chromosome) from your father and 22 autosomes (plus one gender-determinative chromosome) from your mother, for a total of 23 pairs of chromosomes. So this is a test that works across genders to locate relatives — cousins — from all parts of your family tree.
Good: Fewer, Better Matches
The good news, of course, is the big change in the matching algorithm that deletes a very large number of people from all of our match lists who really aren’t related to us genetically at all.
This means that our numbers of matches have gone down — my own results, as you can see here, dropped like a rock: more than two-third of my matches dropped off my list, from roughly 13,000 matches down to about 3,850.
But at the same time, the quality of our matches has gone up.
AncestryDNA’s new matching system begins with a better, deeper, more accurate analysis of the data that helps define who is and who really isn’t genetically related.
One part of the new analysis is through the identification of some parts of our genetic code that really don’t mean anything at all. Some pieces of DNA that we once thought meant we were genetic cousins we now understand just mean we’re all human, or all Scandinavian, or all African. By taking those pieces out of the matching system, AncestryDNA will eliminate many of the false positives.
The other big part of the new analysis is in the way AncestryDNA looks at the bits and pieces of our genetic code that we use for genetic genealogy testing. Since it doesn’t look at all of our DNA, but only parts of it, autosomal DNA testing relies on making some educated guesses about the parts of it the test itself doesn’t look at.
In a way, it’s like reading a book with only some of the words showing up on the page; the analysis system has to guess at what the missing words are. The better the guesses are, the better the results are. So part of the new system is a better way of thinking about what the missing words are likely to be — and in what language. If I’m 100% European, for example, the system shouldn’t conclude that the missing DNA words are in an Asian language.
Emphasizing the right missing words means the people who show up on our lists as matches will be better matches. More accurate. More likely to really be our cousins.
And people we didn’t match before, but who are our real cousins, will show up on our match lists… like a Cottrell cousin of mine who wasn’t on my old list and is on my new one.
Now before people get all bent out of shape about the matches that have disappeared… they haven’t completely disappeared. At least not yet. You can access the old list this way:
1. Go to your DNA Home Page.
2. Under your name, find the word Settings with the gear icon, and click on that.
3. On the right hand side of the settings page, under Actions, there’s a link to “Download v1 DNA Matches” that says we can: “Download a list of your previous “v1″ matching results (available for a limited time).”
We don’t know how long that’ll be there, so if you have earlier matches you want to preserve, download that old match list. It comes in a CSV file format and preserves all of the prior information: if it had a shaky leaf hint, for example, there will be a YES in the HINT column, and if you created a note to yourself about the match, what you noted is reproduced in the NOTE column.
Good and Bad: DNA Circles
The next big change to the matching system is the inclusion of DNA Circles. Think of them as shaky leaf hints on steroids. And they are both good news and bad news all wrapped up in one.
What the circles do is group people who have tested based on their DNA and on their online family trees. Everyone included in a circle will be a DNA match to at least one other person in the circle and everyone in the circle will have a direct line path to some shared person in their family trees. Here, for example, is my DNA Circle for my third great grandfather Martin Baker:
(And please… please, people… respect the privacy of others who have tested. DO NOT go reproducing information without the permission of others who have tested. It’s really easy to black out names and blur photos so you’re not putting information into blog posts or other public places without permission!)
The good part of the DNA Circle grouping is that I don’t have to hunt through every one of my matches to see who else has tested who also lists Martin as an ancestor. This will do it for me. We can now all communicate with each other and share research to see what information one of us might have that the others don’t.
The bad part of the DNA Circle grouping is that not everybody who’s tested with AncestryDNA will show up in your DNA Circles even if you are a DNA match and your trees match. To have this feature at all, you have to be a paying Ancestry subscriber. None of the free accounts will see this. And you have to have a public family tree. Folks with private trees can’t be included.
How deep the family tree data is will affect whether a match lands in a circle too: if the family tree data is fairly complete and the matching algorithm can limit the chances that the match is really in a different line of your family tree, the match is more likely to show up in a circle than if the data is less complete. There’s a lot of information about the circle system and how it works in a white paper that Ancestry subscribers who are in a circle can access (click on the question mark icon when you’re in the DNA Circles area then choose DNA Circles White Paper); at some point, we hope AncestryDNA will make it available to others as well.
And of course it’s still a bad part of this whole thing that we’re forced to deal with someone else’s analysis of what our data really means instead of getting useful tools for comparing the data ourselves.
The Ugly: Misunderstanding DNA Circles
It’s that last point above — that we still have to rely on what someone else is saying about our DNA and our matches — that creates the really ugly part of this change: it’s got a very serious potential to reinforce some very very bad genealogy. And no matter how many times responsible genealogists warn and even AncestryDNA itself warns that DNA Circles do not prove descent from the person identified as the possible common ancestor, people still already believe these circles are proof.
Read through the comments posted yesterday on Facebook and Google+. “I now see who the common ancestor is!” is a common type of comment.
No. No, no, no, no. A thousand times, no.
All we’re seeing is that we are cousins, yes, and that we have somebody that we share in our online family trees. What it does not do is prove that we’re all descended from that person. We could be cousins in an entirely different line — our family tree data could simply be dead wrong.
The whole DNA Circles concept depends in large measure on the accuracy of online Ancestry family trees. And how many times have we all seen major problems with these trees? Someone begins by posting a very complete — but entirely erroneous — line in a family tree, and 10 other people copy it in their own family trees.
Even today, on Ancestry, there are dozens of family trees identifying Samuel Baker and Eleanor Winslow of Massachusetts as the parents of William Baker of Virginia, and Alexander Baker of Boston as Samuel’s father. Um… nope. YDNA testing has definitively disproved that whole notion… but the trees that say it persist and even multiply.
Now if enough descendants get DNA tested and they persist in identifying the wrong people as William’s parents, guess what? We’re going to get a DNA Circle eventually that suggests that they really were William’s parents. And the people who want to believe that they were William’s parents are going to say that the fact that they’re in a DNA Circle with everyone else who claims descent from these same people proves it.
It’s hard enough to convince people to give up cherished notions of common ancestors when the facts don’t support the family stories. But when people are grouped into circles who are in fact related by DNA, getting them to understand that how they’re related may be in an entirely different way, through entirely different ancestors, that the circles are just hints, well… let’s just say I’m not looking forward to this.
Good. Bad. And ugly.
Get used to it.