Friday, June 7, 2013

DNA and Adoption

Last year I used the new autosomal DNA testing at for someone who was adopted.  (Happy Birthday M.S.)  I was eager to compare the services of AncestryDNA to the other companies I use, FamilyTreeDNA and 23andMe.  I was disappointed in the results at AncestryDNA because no comparison utilities were offered.  The site will compare family trees for you and identify possible ancestors in common, but I could not use this function because adopted people cannot create a family tree.

For some background, M.S. was born in 1936 in New Jersey and adopted within the first year of her life.  She knew her birth name, which enabled me to find the court records at the Office of the Surrogate of Essex County, New Jersey.  The papers contained her full birth name, her birth mother's name, and her birth mother's address.  Unfortunately the age of the birth mother was omitted.  [Note:  In 1940, New Jersey sealed adoption records, but not retroactively.  This is how I was able to walk into the records room and read this coveted information.  Most adopted people do not have this option.]

Back in the day (and it's not even that long ago), before so much was online, indexed, and just a few mouse clicks away, I had to manually search papers and microfilm for information on my German families in Newark.  The family of M.S. was of German origin and lived in Newark.  They were usually found in close proximity to my German ancestors, so gathering everyone's information was not difficult.  The paths of our ancestors literally crossed many times in the same neighborhoods, places of work and worship, with the same midwives delivering each new generation.  How ironic that our ancestors most likely knew one another and now our paths had crossed years later!

In spite of all this research, I have been unable to determine which woman in this family was the birth mother.    The older generation has passed, so nobody alive today is able to confirm parentage for us.

A DNA test can connect an adopted person to her genetic family.  At AncestryDNA, the closest relatives for M.S. were at best third cousins.  None had known lines from Germany or New Jersey.

Recently AncestryDNA enabled participants to download their raw DNA data, which had been offered for years at 23andMe and FamilyTreeDNA.  GedMatch is a free site with useful functions for your genetic genealogy studies.  GedMatch expanded to accept and compute AncestryDNA files.  I had already uploaded the raw DNA of myself and others tested at 23andMe.  I was finally able to compare my family to the DNA of M.S.  We *might* match!

I write *might* because the shared segments are smaller than the minimum length generally held to signify a common ancestor [5 cM for the genetically inclined].  GedMatch gave a predicted relationship of about 7 generations back to a common ancestor for M.S. and my father and all three of his siblings.  The results are below.  The areas marked in green are identical areas for the siblings.
M.S. versus my father.
The green area is identical in his siblings and M.S.
M.S. versus my uncle.
The green area is identical in his siblings and M.S.
M.S. versus my aunt.
The green area is shared with her siblings and M.S.
M.S. versus my uncle.
The green area is identical in his siblings and M.S.
The orange area is possibly shared with a cousin who is not related to the suspected common German lines.

In the final comparison above of my uncle to M.S., I have marked an area of chromosome 17 in orange.  This is a possible shared overlap with my father's mother's cousins.  I have not tested the DNA of any cousins from my father's German branch, which is on his father's side, not his mother's side.
M.S. versus cousin 1
The area in orange might be an area where M.S. matches this cousin and my uncle.
M.S. versus cousin 2

If M.S. matches the cousins from my father's mother's side, then this throws my German theory out because the only German in my father's family tree is from his father's father.  I think there is a better explanation.  The area of Chromosome 17 that is identical is perhaps too tiny to be significant.  These cousins match me through their father's side.  If we look at their mother's tree, which is not related to me, we find German roots quickly.

1870 United States Federal Census
New York City
Wrage household

We have another Herman the German.  Herman Wrage was the great grandfather of my two cousins who match M.S.  It is possible that they are related to M.S. as is my father, but on different German lines.  We just have to figure out those lines.

Wednesday, June 5, 2013

DNA Studies: Children of Cousins

In the previous post, I compared the amount of shared DNA between siblings.  Here, we compare the widening differences in the amount of shared DNA between cousins.

With each generation, half the DNA of the previous generation is lost.  Each parent contributes half their DNA to create a new human being.  Beginning with your grandparents, you will not inherit your DNA in precise halves.  In other words, you have 50% of your mother's DNA and 50% of your father's DNA, but not necessarily 25% from each grandparent.  The amount skews with each new child and subsequent generation.

Amount of shared autosomal DNA expected between close relatives.  The actual amounts will vary.

I am fortunate to have been able to test cousins of my grandparents.  My paternal grandfather, Clifford Lutter, only had one first cousin through his mother.  My paternal grandfather, Beulah Cook, had two first cousins through her mother.  If my grandparents were alive to test their DNA, we would expect them to share about 12.5% of their DNA with their first cousins.  Instead, the children of my grandparents (my father and his siblings) have tested their DNA.  We would expect to lose half the DNA with the next generation, placing us around 6.25% shared DNA between first cousins, once removed.  My sister and I are the next generation, so we would expect to share about 3.125% with our first cousins, twice removed.

The comparisons at 23andMe illustrate that even close relatives will share varying amounts of DNA.  Long segments of DNA are broken down in recombination, so that little DNA is identical by the time we reach the third cousin relation.

My grandfather's first cousin versus the next two generations.
The expected amount of identical DNA between first cousins once removed is 6.25%
The numbers for this group range from 6.71% to 9.25%
The expected amount for the next generation is 3.125%.  My sister matches only 2.7% while I match 4.44%

Grandmother's maternal first cousins versus the next generation.
In this set, the actual shared amount varies more widely.
The first cousins once removed share between 3.85% and 6.3%
In the next generation, my sister shares a greater percentage than our aunt and uncles,
while I share far less at 2.14%

Grandmother's maternal first cousin versus the next generation.
The amount of shared DNA among the first cousins once removed ranges from 5.64% to 8.6%
Again, my sister shares more DNA with this cousin than three members of the prior generation,
while I share less than expected at 2.95%


You see in the above comparisons that your ancestor's DNA quickly breaks up into smaller, non-identifiable segments randomly allocated down different descendant lines.  At this point, I have tested most of the known close relatives of my paternal grandparents (except for the descendants of Ruben Charles Cook and Eleanor Lovelace- hint hint).  It is not surprising that most relatives inthe DNA databases share only one tiny segment with me.

1920 Federal Census
Ruben Charles Cook and Eleanor Lovelace
Newark, Essex County, New Jersey