Transfers and interfaces

So there were two big topics of discussion dealing with DNA at the Southern California Genealogical Society Jamboree that ends today in Burbank. One was changes — updates, interface upgrades and promises of further changes — at AncestryDNA, and the other was good news for people tired of waiting for updates, upgrades and changes at AncestryDNA.

These all have to do with autosomal DNA testing — the kind of testing that works across genders and helps you find cousins to work with to fill in information about your common family tree.¹

FTDNA now takes AncestryDNA transfers

The immediate news of use to everyone who has tested with AncestryDNA is that Family Tree DNA can now accepted transfers of raw data from the AncestryDNA testing — and that the price is a phenomenal $49 for one week only, through June 14. Even better, that price, announced at Jamboree, also applies to transfers from 23andMe.

Now let’s be clear: when I say transfer, I don’t mean you’re taking your information away from AncestryDNA or 23andMe if that’s where you tested first. You’re not transferring it in that sense. What you’re doing is downloading a copy of your raw data from the company where you tested and uploading it to another company to be included in its database.

What that gives you, first and foremost, is a better chance at more and better matches. If you’re an adoptee or someone with a lot of brick walls, you don’t want to miss a critical match at company A just because you’ve tested with company B. Just as it’s random chance whether you and a cousin share enough DNA in common to show up as a match, it may also be random chance whether you test at the same company.

So you’re simply taking the test results — a simple text file full of your genetic code data — and adding them to another database so you can fish for cousins in another pond.

What it also gives you, if you’ve only tested with AncestryDNA, is a vastly better interface right now. We’ll talk about the changes AncestryDNA is about to make to its interface in a minute, but even with those changes you’ll still find more and better tools to work with at Family Tree DNA today.

At Family Tree DNA, there are almost unlimited ways to sort through your matches. You can sort by the suggested relationship, the total amount of DNA you and a match have in common, the length of the longest unbroken segment of DNA you and a match have in common, whether you and a match both have a common surname reported in your profiles.

But perhaps the most useful sorting is the ability to identify, quickly and easily, every person you and one of your matches both match: the people you both match in common. And, on occasion, you may want to know who the matches are that you have but that another match of yours does not share. Both of those are options with Family Tree DNA.

There are other tools available in the Family Tree DNA interface that you may find useful, including the ability to see where on which chromosomes the segments are that you and a match share. And there’s the much easier system for contacting your matches: you get the email addresses of your matches (or, in some cases, those who administer the results for one or more matches).

Even if the interface options don’t float your boat, just the ability to see who you match who may only have tested at Family Tree DNA is well worth the $49.

And no, before you ask, I don’t get a thing from Family Tree DNA for saying so.

Changes at AncestryDNA

At Jamboree, AncestryDNA demonstrated interface changes it plans to roll out in the very near future. Once those changes go live, you won’t need to use a third party tool any more² merely to find matches where you and the match both have a common surname in your family tree.

By simply entering a surname into the name field of the soon-to-be-coming search box, you will be able to see only those people you match you have that surname in their trees. And by entering a place into a location search field, you’ll be able to see only those people whose family tree information includes someone born in that location.

The interface retains the sort options available now — sorting by relationship, by the date the match results were posted, and by whether there is a shaky-leaf hint because someone in your online family tree matches someone in your match’s online family tree.

But the interface will not include one option available now: the slider that lets you eliminate from the results you’re looking at all of those real outlier results reported as distant cousins using confidence levels called “moderate confidence,” “low confidence” and “very low confidence.”

AncestryDNA’s vice president for genomics and bioinformatics, Dr. Catherine Ball, noted that there is a very significant chance — stated candidly and clearly by AncestryDNA on the results page — that these matches are false positives — people who actually are not related to you at all. She added that AncestryDNA had chosen to report them as possibilities to give its customers as much data as possible and as many potential cousins as possible to work with.

AncestryDNA has not disclosed how it designates matches as moderate, low or very low confidence, and many genetic genealogists have appreciated the ability to simply turn off those hundreds and often thousands of matches-that-may-not-be-matches-at-all. It would be very useful to have the slider return, both to limit the number of overall matches we have to deal with at a time and to quickly isolate those with common surnames or birth locations who might be the best candidates for a real match.

AncestryDNA’s Dr. Ken Chahine and Dr. Ball both said they understood the desire for more and better tools to work with the underlying raw data but did not think adding a chromosome browser was the way they wanted to proceed. Essentially, they said Ancestry was working on what we might think of as building a better DNA mousetrap. Neither gave real hints as to what form that might take or the time frame in which it might occur.

And they said AncestryDNA would be rolling out a new refinement or revision of the admixture analysis later in 2013. It’s expected to address (and to revise downwards) the unusually high percentages of Scandinavian ancestry reported among people with British Isles ancestors.