Date: Wed,  9 Nov 2005 09:16:05 -0330 (NST)
From: dapike@math.mun.ca (David Pike)
To: PIKE-DNA-L@rootsweb.com
Subject: [PIKE-DNA] Notes from the FTDNA Conference



Hi everybody.

I got back home the other day after attending FamilyTreeDNA's conference
for project administrators in Washington DC over the weekend.  It was a
good conference with lots of information.  Much of it seemed to focus
more on really deep ancestry (ie anthropology) rather than on genetic
genealogy, but I found it all pretty interesting.  It was also a great
chance to meet and chat with other project administrators.

I want to mention some of the things that were mentioned at the conference
which I think will be of more general interest.  One of these is that
FamilyTreeDNA is now processing about 1,000 DNA samples per week.  Most
of these are coming in from participants of the Genographic Project that
is being spearheaded by the National Geographic Society.

The massive growth in business that has resulted from the Genographic Project
has forced FamilyTreeDNA and the U of Arizona to expand their lab facilities
and revise their testing practices.  They've pretty much got a whole new system
in place now, compared to what they were doing a year ago.  For instance, test
viles are now bar-coded, and the DNA analysis process is now nearly all
automated.  The new system not only enhances their ability to process more
samples, but it also greatly reduces the chance that an error will be
made in determining somebody's marker values.  On the matter of errors, these
are also kept to a minimum by testing a set of "control" samples along with
each set of real samples... any time that the control samples do not produce
their expected results, then the folks at the lab know that something is amiss
and will re-run the analysis.  They also take action to investigate anytime
that something anomolous happens (such as when a marker produces no value,
or a value that is not often seen in conjunction with the other markers' values).

Still on the quality control topic, anytime that somebody upgrades to get
their test results refined, some of the previously determined markers are
re-tested as part of the process.  This is mostly so that they can be assured
that the correct DNA sample is being analysed ... if the previously tested
markers do not come out the same as before, then they've got a clear sign
that something isn't right.  My overall impression is that the testing
procedures have been well implemented, along with ample quality control
measures.

One particular question I took to the conference had to do with haplogroup I.
In the Pike DNA project, two of our participants are in haplogroup I, and each
has more than the typical number of close matches in the FamilyTreeDNA database
(well, at least when compared to most other people in the Pike project).  Talking
with other project administrators, as well as asking some of the FamilyTreeDNA
people, I still don't have a really clear answer for what's up with these
two Pike results.  One thing that does seem evident is that there's a fair bit
of puzzlement when it comes to haplogroup I.  One piece of good news is that the
numerous samples now being collected through the Genographic Project should lead
to additional research and discovery regarding haplogroup I.  There are also
expected to be some new developments in the realm of determining subgroups in
the coming months (kind of like R1a and R1b are different subgroups of
haplogroup R), which might help to further distinguish between unrelated
individuals in haplogroup I.

One thing that some of us might have already noticed are some features that
have been added to our personal webpages on the FamilyTreeDNA website.  One of
these features is that we can now upload GEDCOM files.  Some of you may notice
that on your personal "Y-DNA Matches" page some of the matching people have
a small square "FT" graphic ... by clicking on one of them, you will be shown
the corresponding person's pedigree chart, with their male ancestral line
highlighted in blue.  If you want to upload your own GEDCOM file so that those
that match with you can view it, just click on the "GEDCOM" tab on your personal
webpage on the FamilyTreeDNA website and then follow the directions there.  I
should mention that people born after 1900 are automatically hidden from
others' view as a privacy measure.

Our "Y-DNA Matches" pages should also have a small square graphic that allows us
to use the "FTDNATiP" utility (although I note that it doesn't seem to be in place
for exact 12-marker matches with people outside of the Pike project... for 25 or
more markers though, it is present).  Basically it's a utility that uses the observed
mutation rates for each individual marker to compute statistical likelihoods of whether
or not two people share a common forefather within certain periods of time.  I've used
this to determine some of the likelihoods that I've mentioned on the "Results" page
for the Pike DNA Project, but it's also available for each person to use when looking
at the close matches that are listed on their personal "Y-DNA Matches" page.

It was reiterated at the conference that the markers on the 3rd and 2nd panels
(ie markers 26-37 and 13-25) mutate faster than those on the 1st panel (ie the first
12 markers).  This is intentional, because the goal is to try to find markers
that mutate at a high enough rate to be able to help distinguish branches in
a large family tree (but not so fast as to not be likely to linger for a few
generations at a time).  They are working on providing even more markers for us
to be able to make use of, but again the trick is to find new markers that have
mutation rates that are "just right" ... too slow means they almost never change
and too many (unrelated) people will have the same values, whereas too fast means
that even closely related people would likely not have the same values.  I didn't
get any solid impression as to how soon or distant these new markers might be
made available.

Still, it was generally stated that more markers is better.  One person in the audience
compared the 3 panels of markers to US ZIP+4 codes ... with a ZIP+4 code like 01234-5678
the "01" might indicate a state, the "234" the city, and the "5678" the neighbourhood,
whereas with DNA, the first panel is usually enough to determine one's haplogroup,
the 2nd panel might get us to a large family tree, and then with all 3 panels we get
the genetic signature for a particular lineage.  It's a nice analogy, though it likely
doesn't work so nicely in each and every case.  For instance, it was was very clearly
stated that if somebody's markers are very close to a particularly common set of markers
(such as the "Western Atlantic Modal Haplotype" in the R1b group), then it really is
important to refine to 25 or (even better) to 37 markers in order to be able to use
the most available data when comparing two people's genetic signatures.  I did ask if
there's a common haplotype within haplogroup I, and was more or less told yes, and that
there might even be multiple very common haplotypes within haplogroup I, but
I didn't really get any more detail than that.  It was suggested that I get
in touch with a certain person who has some experience with haplogroup I,
and I plan to do that in the next while.

There was one presentation by Terry Barton, who is a coordinator for the Barton DNA
Project, which has about 150 participants so far.  The history of the Barton project
is interesting... they first got started as part of a research project at BYU, have
at various times been affiliated with a number of testing companies, and are now
doing their testing with FamilyTreeDNA.  One thing that I'm sure makes the Barton
project so successful is that the Barton Historical Society has made very generous
contributions to the sponsorship fund for the Barton project.  It was an excellent
presentation, and gave me several good ideas that I can incorporate into some
presentations of my own that I'll be giving in the coming months ... I don't think
I've mentioned these yet, so let me note that I'll be giving talks about Genetic
Genealogy in January 2006 (to the Family History Society of Newfoundland & Labrador)
and also in June 2006 (to the Alberta Family Histories Society, located in Calgary).

As for DC, it was my first time visiting Washington.  The day before the
conference started, I had a chance to walk around the Mall and the various
war memorials.  I dropped by the Vietnam memorial thinking that I might
look to see if there are any Pike's listed on it, but discovered that the
names on the memorial are not listed in alphabetical order.

On a happy note, the folks with Air Canada took pity on me on Sunday
and replaced my DC -> Toronto -> Ottawa -> Montreal -> St.John's
itinerary with one that involved 2 flights instead of 4.

- David.