Interpreting DNA results #dna


Sandra Krisch
 

Dear colleagues,

In searching for living relatives of a Holocaust survivor, Lipman
RADZIK, I discovered in JewishGen's archives that a researcher named
Richard RADZIK was seeking information about his great-grandfather
Leon RADZIK's family. All that was known was that Leon had probably
changed his name >from ROZENSZTEJN to RADZIK after reaching America.

Analysis of anecdotal information and available vital records
suggested that Leon RADZIK had most likely begun his life as Berysz
ROZENSZTEJN, the son of Abram and Sura Ester ROZENSZTEJN of Zuromin,
Poland.

To test this hypothesis, DNA samples were collected >from two men.
One was Robert ROSENSTEIN, a great-great-grandson of Litman
ROZENSZTEJN. The other person tested was Richard RADZIK. If our
theory was correct, he too would be a great-great-grandson of Litman
ROZENSZTEJN

The test results are in. They show that Robert was one of 18 people
who exactly matched Richard on the 12-marker test. Robert was the
only person to exactly match Richard on the 25 marker test. Robert
was one of four people to match Richard on 34 out of 37 markers, but
the only one matching all 25 markers on the 25 marker test.

What conclusions do these data allow us to draw regarding whether --
and more importantly when(i.e. how many generations ago)--Robert and
Richard had a common ancestor?

How does the finding on matches on 34 of 37 markers relate to a
finding of 25 of 25 matches? Does it strengthen or weaken the
likelihood that Robert and Richard are both descended >from a common
great-great-grandfather?

We are calling ourselves cousins. What do we know about the
closeness of our relationship? How confident can we be that we are
correct?

Sandra Krisch
Carefree, AZ


Sam Vass <SVass@...>
 

On 2008.07.21, Sandra Krisch <Sandra@...> wrote:

The test results are in. They show that Robert was one of 18 people
who exactly matched Richard on the 12-marker test. Robert was the
only person to exactly match Richard on the 25 marker test. Robert
was one of four people to match Richard on 34 out of 37 markers, but
the only one matching all 25 markers on the 25 marker test.

What conclusions do these data allow us to draw regarding whether --
and more importantly when(i.e. how many generations ago)--Robert and
Richard had a common ancestor?
The combination of the paper records and the yDNA results together
indicates that there is a relatively recent connection. There is no
known way to "prove" the exact relationship or the amount of time.

Sam Vass, Kent, Washington, USA


Steven Bloom
 

On 2008.07.21, Sandra Krisch <Sandra@...> wrote:

The test results are in. They show that Robert was one of 18 people
who exactly matched Richard on the 12-marker test. Robert was the
only person to exactly match Richard on the 25 marker test. Robert
was one of four people to match Richard on 34 out of 37 markers, but
the only one matching all 25 markers on the 25 marker test.

What conclusions do these data allow us to draw regarding whether --
and more importantly when(i.e. how many generations ago)--Robert and
Richard had a common ancestor?
If you did this test with a good company, such as Family Tree DNA,
then every match beyond the twelve marker match is provided with
some information about **roughly** when your most recent common
ancestor (in this case, a male) is likely to have lived. However,
since such information can be a little confusing, I will try to
de-mystify some of this below. But do please look at your own
results on-line to get the details (and ask someone at the company
if you do not understand them).

Based on the 34/37 match (and the others), you can't say a whole
lot, other than the ancestor probably lived more recently than about
400 years ago (or something like that). This is just an estimate
based on modeling of how Y-DNA is expected to mutate >from generation
to generation.

**But** you have more information. If you match on DNA and a
surname, especially a surname that is not likely to have existed
over 200 years ago, I'd put the common ancestor as having been born
probably no later than 1800. There could be a coincidence here --
such as being linked to someone with a common surname, but without
sharing **recent** ancestry. It sounds like you have already done
your best with records, so there probably isn't too much more there.

I don't see exactly why both men would have to be great great
grandson's of Litman ROSENSZTAJN, but its either that or a
generation or two earlier, not too much more.

In summary I'd say that the records and DNA together show that these
folks are probably cousins with a common ancestor who most likely
lived around 1800, and more definitely since 1600 or so, based on
the DNA alone. But you do want to make sure that you aren't making
too many unsupported leaps in making connections between these men.
I wouldn't be able to say much more without seeing your tree and
documents myself.

It sounds like you are onto something though.

Steve Bloom
Central Virginia


Dave Howard
 

On 2008.07.21, Sandra Krisch <Sandra@...> wrote:

The test results are in. They show that Robert was one of 18 people
who exactly matched Richard on the 12-marker test. Robert was the
only person to exactly match Richard on the 25 marker test. Robert
was one of four people to match Richard on 34 out of 37 markers, but
the only one matching all 25 markers on the 25 marker test.

What conclusions do these data allow us to draw regarding whether --
and more importantly when(i.e. how many generations ago)--Robert and
Richard had a common ancestor?
One should also read the fine print on the FTDNATIP Report. If you
have a paper trail that shows you do not share a common ancestor
then the predictions are moved back by that number of generations.

There is a little box to enter the number of generations you can
prove you do not have a common ancestor. The FTDNATIP report
apparently starts with the assumption that two men are not brothers.
But if you do not share grand fathers then you need to put a 2 in
the box. If you know you don't share great grandfathers then you
need to put a 3 in the box. You will find the results a bit
disheartening. The odds do not get better they get worse, i.e. The
number of generations to Most Recent Common Ancestor gets pushed
back for the same odds.

These predictions are a bit like the long term weather forecasts. As
you get closer to the target date the data gets better and better.
The more markers you have tested the more accurate the prediction.
On Monday there may be only a 10% chance of rain on Sunday. But
Saturday the chance may have gone to 80%.

Dave Howard (Nee Horwitz)
GAP for Horowitz and Jewish_Q


elanc@...
 

On 2008.07.21, Sandra Krisch <Sandra@...> wrote:

The test results are in. They show that Robert was one of 18 people
who exactly matched Richard on the 12-marker test. Robert was the
only person to exactly match Richard on the 25 marker test. Robert
was one of four people to match Richard on 34 out of 37 markers, but
the only one matching all 25 markers on the 25 marker test.

What conclusions do these data allow us to draw regarding whether --
and more importantly when(i.e. how many generations ago)--Robert and
Richard had a common ancestor?
Dear Sandra,

The following web page should answer your question directly.
http://www.familytreedna.com/GDRules_37.html
Or look here for a more general tutorial:
http://www.familytreedna.com/dna101.html

The fact that you know there was a 25 of 25 match may affect the
overall probability of a recent common ancestor somewhat, because
different markers have different mutation rates. The best way to
get an answer is to let FTDNA calculate the probability for you. If
you have an account with the Y-DNA results at Family Tree DNA, you
can log in and click on "Y-DNA matches" in the left hand menu. Then
look through the table of matching individuals for your presumed
relative. Click on the little orange and blue box to the right of
the person's name. This will take you to the FTDNATIP Report, which
will tell you the specific probabilities of the two individuals
being related within a given number of generations.

Regards,
Elan Caspi
El Cerrito, CA


Pavel Bernshtam
 

On 2008.07.21, Sandra Krisch <Sandra@...> wrote:

The test results are in. They show that Robert was one of 18 people
who exactly matched Richard on the 12-marker test. Robert was the
only person to exactly match Richard on the 25 marker test. Robert
was one of four people to match Richard on 34 out of 37 markers, but
the only one matching all 25 markers on the 25 marker test.

What conclusions do these data allow us to draw regarding whether --
and more importantly when(i.e. how many generations ago)--Robert and
Richard had a common ancestor?
FTDNA writes that matching 34 out of 37 markers means that the
probability of a common ancestor in up to

4 generations is 4.3%
8 generations is 28.89%
12 generations is 60.38%
16 generations is 82.15%
20 generations is 93.07%
24 generations is 97.59%

As you can see it is a low probability that the common ancestor was
their great-great-grandfather.

But again - it is all about probabilities, not facts

Pavel Bernshtam

[Moderator's Note: Please note that the probabilities computed by
ftDNA take into account which of the 37 markers were the ones that
didn't match. If you have a 34/37 match of your own, you may have
different probabilities listed on your ftDNATiP Report page.
Nevertheless, Pavel's numbers seem more typical of a 33/37
match than a 34/37 match.]