This project is not part of Ancestry.com DNA sales. This project uses http://www.familytreedna.com Web space provided by rootsweb.com, sponsored by ancestry.com. Please read notice in the bottom bar. Advertisements at the top and bottom of the pages are not part of this project, visiting the links helps pay for the website space ~~~~~~~~~~~~ |
Time to the Most Common Recent Ancestor and Mutation Rates By Roberta Estes
(copyright 2008)[1] One
of the most frequently asked questions is how to effectively determine and use
the Time to the Most Common Recent Ancestor (MCRA) and the TIP tool provided by
Family Tree DNA. You
can access your TIP reports by clicking on the “My Matches” line on your
Family Tree DNA personal page. Those
with a similar surname will have a little red box that looks like a pedigree
chart to the right. Click on that
box and it will run the Tip report for you and the individual whose red box you
clicked. The
results will be presented to you in percent of probability and will look like
this:
The
two men in question above do in fact share a common ancestor 8 generations ago. This
tool was created by Family Tree DNA to give individuals a broad brush idea of
when they might related to another participant, or conversely, if they are not
related. Clearly, this is not an
exact science, and furthermore, it relies heavily not only on your own results,
but on statistical calculations of probability.
While these calculations are very useful in population genetics, the
closer one moves in time to the present, the less useful these calculations
become. Why?
Let’s take a look. Let’s
talk first about using mutations and mutation rates to determine the length of
time that two individuals might converge in a common ancestor. Mutation
rates on the average and for various markers vary WIDELY, and I do mean widely.
In one case, I know of a particular family whose mutation rates in 19
generations are 250 times greater than the average mutation rates. I know
of another family whose rates are significantly lower, nearly zero, in about the
same number of generations. Obviously, both of these are the outliers, but
remember that average mutation rates are made up of just these
things.....everyone's rate is put into the pot and they are all added together
and then divided by the total number. So that rate may or may not be
relevant to you or your ancestors. I strongly discourage people from
trying to figure out or correlate a genealogical connection between two
individuals by using these rates or the number of generations given. It's
just not a science refined enough for that yet. The
good news is that when you run the Tip report for 37 markers or higher, Family
Tree allows you to modify the information using genealogical information, if you
have it at your disposal, which will make the numbers more meaningful, but still
nothing approaching what we would like to see. So
the next logical question is why some of your ancestral families (typically
within surname projects) have no mutations and others have several.
Let's look at how averages are constructed. If any given marker is
on the average (remember what an average was from above) going to mutate once in
every 15 generations, that means that some will not mutate at all, some will
mutate more than once, but on the average you will see one mutation in 15
generations for that particular marker. This 15 generation number is just
an example, as each marker has its own mutation rate. So a mutation could
happen in the first generation, or the second, or the third, etc. If on
the average you have 1 mutation every 15 generations, you don't know what
"click the clock is on", you only when you started looking. This
means that the mutation could have happened is the generation just before you
began looking (i.e. your next oldest ancestor whom you have yet to find) and you
won't see another one for 14 generations, or it could mean that you're
"due" and the mutation happens immediately. It can also mean
that you're NOT due, but you can still have a mutation anytime. So,
we have several factors at play here and whether or not we can measure,
understand or change these factors. First,
the calculated average mutation rate for all markers. This number can be
calculated and is often used, but bears little relevance to individual marker or
family mutation rates. Second,
the individual mutation rate for any particular marker. This can be calculated
and is used by Family Tree DNA. However, several unknowns remain, specifically,
whether or not a marker is more likely to mutate in one direction or the other,
or whether markers on the higher or lower end of the spectrum are more likely to
mutate that markers in the middle of their spectrum. Third,
the personal mutation rate of your own family's DNA.
This can be calculated but you’ll need several individuals with solid
genealogy connecting to a proven ancestor to be able to do this.
Family Tree DNA calculations don’t include this information as they
have no information with which to calculate it. Fourth,
the personal mutation rate of any particular marker in your family. Same
commentary as three above. Fifth,
the "click of the clock". This
remains unknown. Sixth,
the mutation rate of the haplogroup. This
variable is undocumented for the most part, but the fact that there is a
variance is accepted in scientific circles.
How much bearing this has on the rate of the click of the clock is at
this time unknown, and the degree of effect that would come to bear from
variations in the haplogroup mutation rate versus the individual family mutation
rate could be very difficult to determine. Seventh,
generational length, which you could modify somewhat based on any knowledge
about your ancestors and family patterns if known.
For example, in some cultures, generation length average was much shorter
as women began producing as soon as they were physically able.
In other cultures and times, women didn’t marry until they were between
21 and 25, which greatly affects the number of children they can expect to
produce within their lifetime, and the average generational length.
This variable can indeed be tweaked within the Family Tree DNA tip
calculator when evaluating over 37 markers. Eighth,
where the mutations happen to fall in the three panels.
For an example of this, see the final example below. The
best you can do with all of this is to use the tools provided by Family Tree DNA
in their TIP program, but understand just how general these tools are (by the
necessity of the above factors) and only use them as a very general
reference. In particular, do not
run the tip calculator and begin counting up the pedigree chart to the 10th
generation and decide that surely this must be the common ancestor.
You may be right, but chances are, you’ll be wrong.
The TIP calculator and common ancestors are much more reliable when
working with very large timeframes over which to smooth out the effects of the
outliers. For example, in the Estes
example shown initially, the calculator is correct that the two men do share a
common ancestor within 24 generations, and within 20, 16, 12 and 8.
However, that calculation is based on 2 mutations in 37.
If those 2 mutations fell within the first 25 markers, their report would
look like this if run against only 25 markers instead of 37:
If
the mutations for these two gentlemen fell only in the third panel, then they
would match at 100% if they had only tested 25 markers, and then their report
would look like this:
Remember
that this example is using the same number of mutations but placing them in
different panels and running the tip against either the 25 or 37 panel models.
The
moral of this story is that we should use all of the tools available to us, with
our eyes wide open, and always, always evaluate the scientific data in light of
the genealogical information available. However, like most everything else in
life, there is an exception to the rule. If
the DNA says you’re not related, then you’re not related.
I have intentionally selected two men who I know are not related to
demonstrate this phenomenon.
DNA
can disprove genealogy, but genealogy cannot disprove DNA.
However, these statistical tools are not infallible.
Another
handy feature provided by Family Tree DNA is a chart that gives rules of thumb
for matches and mismatches. Generally, I would agree with this, but again,
averages are compiled by including all of the data, including outliers, and this
is not gospel, it is a statistical guideline.
Looking at the following chart provided at http://www.familytreedna.com/gdrules_12.html,
keeping in mind that while we do call 3 mutations in 12 generally
“unrelated”, if these same 3 mutations were the only ones out of 37, we’d
be calling the two people “related”. It is possible, if improbable, that
someone could have 3 mutations in the first panel and match on the rest.
The only way to know is to perform further testing.
For
more detailed information about the TIP calculator, refer to the FAQ page at http://www.familytreedna.com/faqtip.html. Under
what circumstances can we best utilize these tools?
I use these tools most often to determine whether people with differing
surnames are related in a genealogical timeframe, especially if I find them
co-located. Again, TIP is most
accurate the further back in history you can reach.
It is particularly useful in answering questions like are the McMahon
family and the Estes family really descended from a common ancestor and if so,
in what timeframe might we reasonably expect to find that common ancestor.
I also use TIP as a sanity check when the genealogy is questionable.
TIP is useful if we don’t expect it to perform more than it can, don’t use it to mis-set our expectations and or to “confirm” things like “two men share an ancestor within the last 8 generations”. TIP can’t do that, and this type of misconception and over-interpretation is all too common with excited first time participants who have just received their results. Using this tool correctly and understanding the results can indeed help us move our genealogy project forward by accurately utilizing both genetics and statistics. [1] TIP is trademarked by Family Tree DNA. The charts and graphs reproduced here are from the Estes surname project and are copyrighted by Family Tree DNA.
|
Contact Information: Electronic mail General
Information/Project Membership:
[email protected]
NoticeThe Lost Colony Research Group is in NO WAY affiliated
with The Lost Colony Center for Science and Research. The Lost Colony Y-DNA and MT-DNA projects at Family Tree DNA are
NOT IN ANY
WAY affiliated with The Lost Colony Center for Science and Research,
"Please
notify us of any claims to the contrary." There is no fee to join our group and no donation of monies or objects are needed to participate in "The Lost Colony Research Group". ~~~~~~~~~~~~~~~ As with any DNA project, individuals pay for their own DNA testing, but the ~~~~~~~~~~~~~~~ Neither Rootsweb.com, myself, nor the Lost Colony Research Group together or individually are responsible for the personal content submitted by any individual to this website.
Send mail to [email protected] with
questions or comments about this web site.
|