Changing Market for Home DNA Testing

Herbert Lazerow

    Part of the problem with dna testing services is that our expectations are too high and their explanations are too minimal, and often not read.
    1. There is a lot of "noise" dna out there that matches, especially for Ashkenazi Jews and other communities with a long history of intra-marriage.  For that reason, a large number of centiMorgans of match may not tell us much except that at some point in history, there was a common ancestor.  The problem is that for Jews whose ancestors come from eastern Europe, records only began to be kept in the 1800s, and many of those records have since been destroyed or temporarily lost.  For that reason, we can neither confirm nor deny a relationship based on a common ancestor who was born much before 1800.  What tells us in a rough way how long ago the common ancestor lived is the second number, the length of the longest consecutive string of matching dna.  It seems to require a minimum of 20 cMs, and more often 40 cMs, to find a common ancestor after 1800.
    2. We each can be sure that we are receiving 50% of our dna from each parent.  Thus, in a perfect statistical world, grandfather (gf) would have 50% of greatgrandfather’s (ggf) dna and 50% greatgrandmother’s dna.  Gf’s son father (f) would have 25% ggf dna, and father’s son (s) would have 12.5% ggf.  However, that is not exactly how it works because
    3.  the 50% of each of your parent's genes you receive is a RANDOM half.  Assume that ggf had only 8 genes, A-H.  Assume further that gf inherited 4 of them, genes A-D, 50%.  F inherited 50% of gf’s genes, but which ones? It could have been A-D, in which case father would have 50% ggf genes, just like gf. If father inherited genes E-H, father would have no ggf genes. Either result is possible, though neither is the most statistically likely.  It is more likely that f inherited 3, 2, or 1 ggf genes. Let's say he inherited 2, though it might as easily have been 1 or 3.  Now I make my grand entrance into the world.  How many ggf genes do I get? Maximum is 2, because that was all f had.  I am more likely to get 1 than 2, and more likely to get none than 2, because I had only a 1 in 4 chance of getting any, since only 2 of father’s genes were ggf genes.
    4. Which seems depressing for the likelihood of finding relatives by dna testing, but the good news is that we each have more than 8 genes. We have about 3 billion “base pairs”. Even at the remove of 5 generations, there is likely to be some evidence of relatedness. 
    This is why my test reports such different numbers of matched centiMorgans for people who bear the same relationship to me. My known second cousins vary from 291 total matched centiMorgans/75 cMs longest matched segment to 240/43 to 200/21.  I have a third cousin with whom I share only 103 cMs, but our longest consecutive string is 60 cMs.
    Sometimes I think that science is not nearly as precise as we non-scientists believe. Or perhaps it is very precise, but we need to learn to read it better.  For Ashkenazi Jews, big total cMs are not as important as long length of longest strand.
    That raises three problems with home dna testing companies. I have tested with Family Tree DNA and Ancestrydna, and posted my Family Tree results to MyHeritage, so I can only speak about those services.
     The services vary in different ways.  Ancestrydna advertises that it has the largest database.  My hunch is that Family Tree DNA has the largest Jewish database.  MyHeritage and Family Tree DNA provide a host of services and information for comparing your results with others. Ancestrydna does not even provide the length of the longest continuous stretch of matching dna; you can try to figure it out from the number of segments counted and the total amount that matches, but that is just a guess.
     Those services that ask for ancestral surnames do not ask very effectively.  The time to ask is in your application for the testing kit, but that would slow down the moment when the customer pays his money.  They could direct you to an appropriate form when they send you notice that your results are ready, but they do not. You must find it yourself.
     It is incredibly frustrating to have a match and message the person, but receive no response. I always send my ancestral surnames and locations to my matches.  I assume that a person who does not respond finds no matching surname or geography, but that is not always true.  Some known relatives have not responded. An alternate explanation is that these are people who received the test kit as a gift or who did the test as an accommodation to a close family member, but are really not interested.
     On messaging, there is a big difference between Family Tree DNA, which provides the e-mail address of your matches, and Ancestrydna and MyHeritage, which require that you communicate through their internal messaging service.  For a period the latter did not allow use of the messaging service if you did not continue to subscribe.  I think they have discontinued that limitation.
     The changing market for home DNA testing may simply reflect a changing fad, disillusionment due to excessively high expectations, too many companies dividing the market, or the fact that most people who want testing have done it. Company mergers might benefit users because they would increase the size of the database used for comparison, but one must always be aware that there are competition laws both in Europe and the U.S. that may present legal impediments to mergers, or even sharing of databases between companies.
Herbert Lazerow
Professor of Law, University of San Diego
5998 Alcala Park, San Diego CA 92110 U.S.A.
(619)260-4597 office, (858)453-2388 cell, lazer@...
Author: Mastering Art Law (Carolina Academic Press 2015)

Join to automatically receive all group messages.