BARBEE-DNA-L Archives

From:
Subject: New Relationship Calculator at Family Tree DNA
Date: Mon, 27 Dec 2004 18:29:40 EST

Hi everyone,

Hope everyone had a wonderful holiday! We have one more present under the
tree from FTDNA.

I have mentioned before that the University of Arizona Genetics Department
did a study of mutation rates among families who were tested at FTDNA. At the
Group Administrator's conference last October, they announced this and gave
us partial information on their latest estimates of mutation rates.
Basically, their findings are different than those of previous studies - they found
that the mutation rate for all 37 markers is about double of what was
previously thought. They also determined mutation rates for each specific marker.
They will not release the marker-specific mutation rates until they have
published their research in scientific peer-reviewed journals. However, they did
give us some hints:

-- there are 4 markers that mutate much more rapidly than the others. (I
bet we can guess at some of them based on our results!)

--the average mutation rate for the first 12 markers is less than the
average mutation rate for 13-25 which is less than the average mutation rate for
26-37. In other words, within a family, we should see close similarities in
the first 12 markers, slight differences in the next 12 markers, and the most
variation in the final 12 markers. That is exactly what our results look like
for the William Barbee line.

--by knowing the marker-specific mutation rates, better estimates of
distance to the common ancestor is possible.

--the estimated time to the common ancestor will be sooner than previous
estimates because the University of Arizona's research shows that the markers
mutate more rapidly than previously thought.

Group administrators are an impatient bunch and we have been begging FTDNA
to get us the specific mutation rates ASAP. Although it may be a long while
before they publish and release the rates, FTDNA has developed a "calculator"
which they call FTDNATIP that each person who has been tested at FTDNA can
use to calculate "Time to Most Recent Common Ancestor". (This is available to
you only if you have signed the release form). Below are excerpts and
instructions from FTDNA's announcement.

I have used our fearless co-administrator, Greg, as an example and compiled
his FTDNATIP results on a temporary web page so you can see what to expect and
to illustrate a couple of things. You won't find a link to this page yet
on the DNA web site. Eventually, I may try to do a table like this for
everyone. (I said EVENTUALLY!!! It takes a lot of work.)

http://members.aol.com/BarbeeJK/DNA/ftdnatip24644.htm

Here are some things to notice in Greg's example:

--The columns on the right show the probability that Greg and each of the
other participants are related within 100 years, 200 years, 300 years, etc.
FTDNATIP assumes an average of 25 years per generation or 4 generations per 100
years. It is commonly known that there is a lot of variation in the average
number of years per generation among families. Personally, I think this
introduces another chance for imprecision in the estimates, so I have indicated
the number of generations in each column. It seems to me that the number of
generations is more useful in trying to track down that elusive common
ancestor. FTDNA may incorporate this in a future version of FTDNATIP.

-- We know that Greg is related to 23733 and 24809 and we know who the
common ancestors are and how many generations back they are. For example, look
at the row for 23733. Greg and 23733 are 7 and 8 generations away from their
common ancestor and their common ancestor was born about 260 years before
Greg and 23733. They have a genetic distance of 2, which means that they had a
difference of 1 on 2 different markers. (You can look at the Results table
to see this). FTDNATIP calculated that there was a 68.67% probability that
Greg and 23733's most recent common ancestor was born within the last 200 years
(8 generations) and an 89.14% chance that the most recent common ancestor
was born within 300 years (12 generations).

-- Also notice that the probabilities are different even if the genetic
distance is the same. For example, Greg has a genetic distance of 3 mutations
from 22893, 24809 and 24823. But the probabilities are slightly different for
each of these three individuals. The reason for that is because they
differed from Greg on different markers and those markers had different mutation
rates that are used in the FTDNATIP calculator.

-- Notice below the first table, there are some other tables for specific
individuals (those who had a genetic distance from Greg that was greater than
zero). These are fine-tuned estimates (using FTDNATIP) of the distance to the
common ancestor based on our "prior knowledge" that Greg and each of these
individuals could not have possibly been related within so many generations.
Personally, I'm not a statistics whiz so I find some of the results
counter-intuitive. The probabilities seem to be less than they are in the first
table.

If those on this mailing list are interested, I have the University of
Arizona average mutation rates for 1-12, 13-25, 26-37 and 1-37.

Please post any questions, comments or observations to this list so we can
all learn from each other. I have not analyzed anyone's information other
than Greg's, so each of you are likely to notice different things as you play
with this calculator. I thought Greg made an interesting example since we know
he is from the William Barbee line and there are two other known William
Barbee descendants to compare to.

An excerpt of FTDNA's message follows. Note: Follow their directions and
click on the "ancestral tree" symbol to the right of each person's name whom
you match. Also, don't bother running the test for 12 or 25 markers unless
that is all you have tested.

Happy holidays and have fun!

Janet


In a message dated 12/27/2004 1:59:32 PM Central Standard Time,
writes:

To Barbee group manager,

- FTDNATiP is now open to all members of Surname Projects: after the
extremely positive response from our Group Administrators, we have decided
to offer this resource to all our Surname Project members so that any
potentially relevant matches can be analyzed with FTDNATiP. This tool can
be found at the "Y-DNA Matches" section and we encourage our customers to
read the "Understanding FDNATiP" page to learn why some matches have the
FTDNATiP tool for comparison, while others don't.

- E-mailing the FTDNATiP Report: because the Report is located on your
members’ Y-DNA Matches page, it is restricted to the most relevant
matches from within your group. This feature will allow you to make any
comparison between members and e-mail the Report page to the individuals
being compared. The page that is e-mailed is in PDF format for easy
printing.

On a final note, a quick reminder to those of you who still have FTDNA
matching funds ($$) in your General Fund: the use of those matching gifts
will expire on 12/31/04, and after that day we will remove them from the
log, leaving only the actual donations. If you need assistance on how to
make the most out of it, you are welcome to contact us.

Happy New Year!

Max Blankfeld
Vice-President, Operations and Marketing

This thread:

• New Relationship Calculator at Family Tree DNA by