Friday, July 17, 2020
Cheering for the Underdog
After the word got out about AncestryDNA's upcoming plans to adjust matches upward to a cut-off of eight centiMorgans from its current minimum of six, I spent the day reviewing all my tiniest matches for possibilities to save. Among those were connections to surnames such as the Falvey puzzle I'm currently wrestling with, but they also included links to other distant ancestors whose stories I am struggling to preserve—such as the oral history of King Stockton repeated to my grandmother, then my mother, before I ever heard it told. It is so important to keep connected with these other researchers, especially after being promised that the bulk of messages in the old Ancestry messaging system could be downloaded—but yielded nothing, once the downloaded file was opened.
That, however, is another story. For now, the issue is to save what we may regret losing in only another couple weeks. While the task may seem tedious—and enormous—I am taking it day by day, lest it get overwhelming. With a system, I'm marking each six or seven centiMorgan match if it contains a tree with one of my rarer family surnames, or presents with one of those Ancestry green leaf matching ancestor icons. I'm not going so far as to color code them now; without certainty about the validity of the connection, all I want to do is flag them so they aren't lost to future consideration.
Meanwhile, a recently-released white paper helps put the recent changes at Ancestry in a more positive light. For those with a deep need to know more than anyone else would care to know, the thirty-four page document is available here.
With all the awe-inspiring wizardry and computational prowess that goes into determining our personal genetic makeup and who we most closely match, there still is a downside to the methods currently employed at AncestryDNA. Granted, brainiac algorithms have yielded us an amazing body of knowledge, but Ancestry, like the proverbial giant, has difficulty with any move it makes. There are computational costs they are confronting with the largest genetic genealogy database currently in existence, true, but compounding the challenge is the fact that this eighteen million count is not a static number; the database is constantly growing. This, too, brings on its own challenges—and increased costs.
To add to their arsenal of algorithms to produce the results we have come to expect, the AncestryDNA staff have developed additional procedures, such as the genotype phasing strategy they have dubbed "Underdog." Reduced cost and speed of computation are business goals, granted, but the challenge of a constantly-changing data set as it spirals up beyond its current benchmark of eighteen million is likely uppermost in the minds of Ancestry executives.
Much of the outcry over the announced changes may turn out to be needless drama, but we'll have to wait to see how our own matches change—if at all. If I weren't in the throes of wrestling with a reticent brick wall ancestor right now, perhaps I wouldn't have cared at all to see the announcement this week.
Then, too, I need to remember that I also have matches at all the other DNA companies where I and my relatives have tested. Beyond that, as we progress with examining some of our research problems, we can't lose sight of the opportunity to find key relatives to ask to consider adding their test to the DNA mix. There are many other ways to utilize the tools we have at hand, even if the decision at Ancestry turns out to be as drastic as some make it out to be.
In the meantime, just in case, I'll be continuing that plodding system to quickly earmark any possible Falvey connections whose match just missed that magic eight centiMorgan mark. Ancestry may wish to kiss them goodbye, but I'm not quite ready to cut the cord just yet.