Getting closer: Test results on WikiTree.com
Chris Whitten
all your messages. I think we may be getting closer to something.
First off, I want to say that I appreciate this comment from Dr. Bill
the other day: "While I believe the discussion is useful, my first and
second reaction is that, at this time, at least, it is an unnecessary
and undesirable distraction from the important work that WikiTree is
set up to accomplish." (I also appreciate the embedded compliment.)
We do have very limited resources. Since this is designed to be an
entirely free site we need to keep costs low. And this means we need
to be extra vigilant about staying focused on what matters most.
We want to stay away from things that could be done just as well (or
better) by someone else. We only want to undertake things that move us
toward our end goal -- a worldwide family tree -- and are consistent
with our chosen means -- a collaborative wiki environment.
I think integrating DNA results may fit these criteria. How much of a
priority it should be is less certain.
I can say with near-certainty that any significant technical changes
we decide on won't be ready for at least a couple months. There are
too many other high-priority features and functions in the works, and
there are always unexpected things that come up. Today our resources
were consumed with foiling a new spambot. The bastards.
Anyway, I see three ideas gelling:
1.) Ability to attach results to a profile. We want these to be in a
consistent, universal format so that results from different testing
agencies can be compared with each other. I'm still not entirely sure
the format is universal. Can someone confirm this?
2.) Linking to surname groups here or elsewhere. There's no scientific
link between surnames and DNA, of course, but there's an interesting
connection to be made. Where would these links go on WikiTree? We do
have surname pages -- e.g. http://www.wikitree.com/genealogy/Hunter --
that might be used somehow.
3.) Automatic searching for matches. If the results being attached to
profiles (#1) are stored in a universal format, and it's text not an
image, they can be matched fairly easily as John describes.
Here's one thing that concerns me. Does anybody understand Family Tree
DNA's new "Family Finder" test?
http://www.familytreedna.com/landing/family-finder.aspx? They seem to
be claiming that this doesn't use yDNA or mtDNA. If that's true, and
this test became popular, would anything that we're doing with yDNA
become less relevant?
Thanks again for all the great input. This discussion is really valuable.
Chris
--
Chris Whitten
Creator of WikiTree.com
http://www.wikitree.com/wiki/Whitten-1
Val Van Zee
Not sure I can say exactly that I "understand" Family Finder (FF), but have
been experimenting with how it works for almost a year now. The way FF is
set up now, I can't foresee that it will ever make y-dna less relevant.
There has always been an interest in one-name studies and y testing is
really the only dna product to compliment those types of projects. But, of
course, that's only the tip of the iceberg. FF will be helping us address
all those other brick walls we all have regarding our grandmother's father
or our maternal great grandfather's mother's line. The raw data which backs
up this kind of information is huge for just one single individual. I don't
yet have any kind of a conception as to how this kind of data can even be
discussed within any kind of narrative about a certain individual...
hopefully that part will become more clear with time.
You might notice I qualified the above statement with "the way FF is set up
now", because they have just changed the platform they use to analysis
these results. One of the reasons they give for making this change is to
make it easier to integrate upgrades. I have no clue about what may be in
store for us as autosomal DNA advances.
Regarding y-dna, if you are really going to try to work some kind of
database into WikiTree to compliment profiles, you best be prepared to be
able to handle strings of numbers at least 178 boxes long. FTDNA recently
started offering an 111 marker upgrade. The current challenges of showing 67
boxes already creates a good deal of havoc for some websites... a number of
smaller (more family focused) website presently only make 25 or 37 marker
available for viewing. I can't even imagine how such lengthy strings will
be handled on FTDNA and/or WorldFamilies websites which are the primary
reporters of such results. Of our 150+ members only one person has ordered
this upgrade to date, but I'm sure there will be more to follow. Thus far I
have received no administrative suggestions as to what changes will need to
be made to the Gibson Results Page to accommodate these lengthy strings.
I'm unclear why you are quite so pro-active about WikiTree incorporating
this kind of a database on a site you describe as space sensitive. Though I
will add here: there does seem to be needed some kind of industry standard
for reporting results. There is quite a difference between the way various
sites report results... even within FTDNA surname projects, administrators
seem to have a wide latitude as to exactly how their results are reported,
also true for WorldFamilies sites. Then there are the "family" sites which
also report results the way they deem makes the most sense for them. In
spite of these inconsistencies, databases already exist and if a submitter
of a WikiTree profile thinks it's appropriate they can create a link to such
sites and make their comments within the WikiTree profile. DNA is still in
it's infancy. I don't pretend to know what might be develop next year, much
less 5 years down the road, but l will predict that there are still huge
changing in DNA looming in the future.
Val
Val Van Zee
I missed addressing one of your primary questions and one of your comments:
> Ability to attach results to a profile. We want these to be in a
> consistent, universal format so that results from different testing
> agencies can be compared with each other. I'm still not entirely sure
> the format is universal. Can someone confirm this?
Comparing results from different labs is not totally compatible today.
Companies do not test exactly the same markers, so there is no "universal"
standard at this time.
> Automatic searching for matches. If the results being attached to
> profiles (#1) are stored in a universal format, and it's text not an
> image, they can be matched fairly easily as John describes.
If everyone (most everyone) who has tested or will test at the 67 marker
level will also create a WikiTree profile we could have a meaningful
database.... a desirable goal, but perhaps not totally realistic.??? It also
seems to assume that 67 markers will forevermore be the "industry standard".
Another caveat: remember such tests and test results are not "public domain"
(as with recorded instruments such as deeds, wills, etc), they belong to the
certain individual who submitted the sample... or perhaps in a small
percentage of cases to the individual who paid for the test, if not the same
person.
Val