For quite some time now, the genetic genealogy community has been beating the living tar out of Ancestry.com for not listening, among other things. Well, I’m here to say, they are listening. Now, what I can’t say is how much they are hearing. The jury is out and we will see. However, we are hopeful.
Ancestry invited a few of the leaders in the genetic genealogy field to come and meet with them this week. They dedicated the resources of eighteen of their scientists and executives to this meeting and they spent the day with us, sharing information about the science underlying their upcoming product changes and having frank discussions with the group.
This was a very cordial, informative and I think, team-building, experience, but there was far from uniform agreement. There was a great deal of discussion which I think helps everyone understand the position and reasoning of the other parties involved. Like anything else, it’s not as simple as one might hope.
Another important aspect of these meeting is that they serve to put faces with names and humanize the other people involved.
I also found it encouraging that most of the people at Ancestry are genealogists and utilize their own tools.
Tim Sullivan, CEO of Ancestry stopped by and talked with us for a few minutes. He asked us what we wanted, why and if we had any questions for him. He told us about his own genealogy experiences. And, we discovered, he does read our blogs. Tim is very actively engaged as is Ken Chahine, Senior Vice President and General Manager DNA, who is in many of the photographs because he was sitting at the end of the screen and was with us for the entire day.
I will be covering different aspects of the content of these meetings as time moves forward and as Ancestry’s new software version is implemented, but for now, I wanted to update you on the two burning questions in the genetic genealogy community.
These, as you might guess, were also the most contentious aspects of the entire meeting.
Will We Receive a Chromosome Brower?
I want to share with you readers that there is absolutely no question that Ancestry heard the message that we need a chromosome browser, loud, clear and uniformly from us. Ancestry is equally as adamant, it appears, as we are, that we don’t need one.
So, the short answer is no.
The longer answer is probably not.
Judy Russell, in comments to her article, “when less is more,” which I strongly encourage you to read, says about the chromosome browser:
“In my personal opinion, speaking only for myself and based solely on my own perceptions of the attitudes of some folks at AncestryDNA and not on any specific representations by anyone else, my judgment is that we may get a chromosome browser at AncestryDNA when hell freezes over.”
I think this pretty well sums it up.
I do want to discuss why Ancestry does not feel a chromosome browser is warranted. This topic was discussed directly and indirectly several times throughout the day. These concerns listed below are not necessarily in priority order based on discussions, because I couldn’t really discern a priority.
1. Given that Ancestry will hit the million kit DNA mark sometime in the first quarter of 2015, they feel that very few, a small percentage, of those people would ever utilize, or understand the results of a chromosome browser. Given that, they don’t feel it is a good investment of their engineering time to invest in something that few people, or a small percentage of the whole, will utilize.
2. Since Ancestry did not begin utilizing chromosome browsing in the beginning, they are concerned about privacy issues having to do with now introducing the feature to people who did not expect to have that to begin with.
3. Ancestry is concerned about unexpectedly and unintentionally revealing health information. For example, let’s say that today, a particular SNP is included in their information and is not known to be medically relevant. Next year, someone discovers that a particular SNP on chromosome 7 is connected to the genetic propensity for erectile dysfunction. Remember, a genetic propensity does NOT mean you have or will get the particular disease. In this case, of course, that would not apply to women.
Ancestry’s concern is that since they would have already been displaying that match on chromosome 7 between several people for months/years, the cow is proverbially out of the barn and closing the door at that point it a bit late, if possible at all.
Of course, as we pointed out to Ancestry, that’s the entire point of having testers sign a release, and both Family Tree DNA and 23andMe both deal with the same issue.
4. Ancestry feels that a chromosome browser would provide information to people that they should not be drawing conclusions from, and they are.
For example, as they showed us, there are areas in each person’s chromosome and their matches chromosomes that are what they call “pile up” areas. These are areas that we would call IBS, identical by state as opposed to IBD, identical by descent. Some of these pileup areas are so old that they could potentially be considered AIMs, or Ancestrally Informative Markers that harken back to continents like Africa, Asia or Europe.
This slide shows Cathy Ball, VP Genomics and Bioinformatics, showing me my own pileup areas. The two screens are a TV screen to the right where the colors resolved much better, and the larger screen where the display was larger.
What this shows you is that on the chart at left, I have one area that has a very large number of pileups, probably about 800 matches (out of my 12,500 total matches), two areas that have 400 each, two that have about 200. On the chart at right, the top of the chart is 25 match segments, so you can see that most of my matches fall below that. Ancestry feels that the higher matched segments are less relevant because they match to so many people, that they aren’t really indicative of shared ancestry in a genealogical timeframe.
And no, they did not tell me which chromosome these pileup segments are found on, and I’m DYING to know so that I can relate that to my ancestral chromosome mapping….but no cigar. It’s so frustrating that they know, they have the info, our info, but they won’t share it with us. I’m not referring here to the slide and my pileup, but the lack of segment information in general. I don’t know how that’s any worse that allowing customers to infer that a shakey leaf tree match is equivalent to a DNA match…..
Everyone has these pileup areas, which also means that they show up on your chromosome browser as matches. Ancestry is concerned that you will see three people, whether from a common genealogy line or not, who match on one segment and you will presume that they are genealogically related, when perhaps they aren’t, because their match is IBS from a pileup area.
Clearly, those of us who work in this field daily deal with IBS issues routinely, but Ancestry is concerned about the general consumer who doesn’t.
I suggested that the chromosome browser could be even more useful if they had a way to show but “grey out” those pileup areas, so we would be aware that their confidence is low, and to highlight the areas where the rarest alleles match, because those matches are most likely to indicate true genealogical matches. That suggestion met with polite silence.
I do agree that many people won’t utilize the chromosome browser, but many people won’t utilize many of their services. That doesn’t prevent Ancestry from providing those services for those who want to utilize them. I’m fine with Ancestry making the Chromosome Browser part of a subscription kit so only subscribers have access, just like many of their data bases.
Unfortunately, without a chromosome browser, we are left with nothing concrete to base any matches on, nor the ability to utilize that information in conjunction with chromosome segment information from other companies to map our segments to various ancestors. The problem of incorrect ancestor attribution remains and will remain present in their matches.
They are changing their matching algorithm and in some ways, it will be improved, but in one way, I am gravely concerned that it will be worse. Ancestry will begin weighting various factors in calculating the match strength, and one of those factors will be the number of trees that list a particular ancestor. If you’ve just had a coronary…so did we. I thought one of the genetic genealogists was going to have the big one right there – they turned so red in the face.
A second confidence weighting factor will be the amount of source information for a particular tree which Ancestry feels helps judge the quality of the tree. In a sense, I agree, but attaching source information, perhaps incorrectly, to the wrong family, or having the wrong ancestor you’ve just attached source information to, is still the same large problem. Clearly, quality is not a matter of quantity, but just as clearly Ancestry cannot look at each tree individually and render an opinion, so they have to develop some automated methodology if they are going down this path.
Ancestry is trying to find ways to improve their matching and predictions of common ancestry. As time moves forward, I’ll be covering these developments. As someone in the meeting said, first steps first.
But back to the chromosome browser, my gut reaction to this is, and this is my opinion alone, that they don’t want to invest the development effort into something that will make the user experience more complex and may increase their customer support staff load to support and explain matching on a chromosome browser. I don’t think they believe the genealogy community has the ability to utilize and understand this type of tool. Ancestry is a genealogy marketing company. They want the user’s experience to be pleasant, easy and fulfilling…not difficult and certainly not upsetting.
Our message did not waiver, we need a chromosome browser and “trust me” simply won’t work.
The Y DNA and mtDNA Data Base
When Ancestry sent the invitation to this meeting, I had to wonder if they really thought through the fact that this meeting would occur less than a week after they decommissioned their Y and mtDNA data base.
Did they really want a group of people that were mad as wet hens arriving to meet with them? I fully expected to receive an “un-invitation” after my article and before the meeting, but I didn’t.
Without going into nitty-gritty detail, Ancestry indicates that the data base that held those results was literally on its last leg and they did not want to invest any money into something they was not bringing in any revenue and for a product they were no longer selling. I do believe that data base was indeed in its death throes because after the denial of service attack in June, it was no longer searchable.
In the ensuing discussion, the genetic genealogy community provided a number of alternative scenarios both within and outside of Ancestry as a way to salvage the information in that database. Ancestry has agreed to take the matter under consideration internally and discuss the various options. They made no promises, but I personally find it very encouraging that they are willing to discuss the matter and reconsider.
I told them I’d like nothing more than to write a retraction article that says that Ancestry did not, after all, burn the DNA courthouse.
In the same vein, I asked if they had any plans to decommission the Sorenson data base at www.smgf.org and they indicated that they do not have any plans at this point to do that. Obviously, nothing is forever, and they could reconsider in the future but at least it appears that resource is safe for now and adding the Y and mtDNA records from Ancestry into that data base was one option discussed.
I do feel this was a productive meeting. The scientific aspects of having a large data base to draw from are quite interesting and I’ll be sharing some those in upcoming articles. Some of the best conversations took place beside the proverbial “water cooler.” I am hopeful that we made progress, or at least thawed the ice a little on the issues so critical for the genetic genealogy community, but time will tell. In a way, I felt like this was a United Nations type of meeting where everyone leaves with a better understanding.