Which DNA test to buy for genealogy?

People ask me often:  Which kind of DNA test to buy?

I tell people they should do an autosomal DNA test.  This is the only test sold by AncestryDNA and 23andMe.  Family Tree DNA also offers this test, called Family Finder.  The price has dropped to around $100 at all three companies.

An autosomal DNA test matches you with other people whose DNA is identical to yours in a few spots, as most will be distant cousins.  These identical areas were passed down from an ancestor common to you and your match.

Diagram of Autosomal DNA inheritance

The goal is to figure out which of your ancestors you have in common.  This is not easy.  The DNA test will not produce a family tree for you.  You still must research your ancestors in records to find the leaves of your family tree.  You may find a close cousin who has done a lot of research and this cousin may provide you with his/her research.  You may find others who help you work past a tail end in your tree suspected of holding the ancestor in common.  If you were adopted, you may find close relatives who can help you identify your biological parents.

If you are serious about finding relatives, you need to test your DNA at all three companies.  You may have a close relative at one site who is not going to test multiple times to find other relatives.  You have to find them by testing at all three sites.

Confusion arises with the Y-DNA and mtDNA tests.  These tests are offered by FamilyTreeDNA and vary in price based on number of tested markers.

One ancestral line only:  Y-DNA testing and mtDNA testing

The Y-DNA test is for males.  Testers are matched to other men who share an almost identical Y chromosome.  The value of this test is that you don't have to figure out which ancestral line holds the common ancestor.  The common ancestor will always be on the direct paternal line, father to son, because this is how the Y chromosome is inherited.  Fewer differences with a match indicate that the common ancestor is within fewer generations.  This test is useful in surname studies, if your direct paternal line comes from a community that perpetuates last names from father to children.

The mtDNA test (mitochondrial, not maternal) is similar to the Y-DNA test in that the common ancestor will be in one ancestral line, mother to daughter.  MtDNA is passed on from mother to children, both sons and daughters.  Men and women may take this test.  A man will not pass his mtDNA on to his children, so that is why the test-taker can be male, but the common ancestor will be found in his mother's direct maternal line.  Unlike the Y-DNA test, the most recent common ancestor could be thousands of years ago, as mtDNA does not vary generation to generation as much as Y-DNA.

These Y-DNA and mtDNA tests provide your haplogroups, which is useful in tracing ancient migratory paths of humans across the globe.  23andMe provides predicted haplogroups with their autosomal test.  This is not the same as taking a Y-DNA or mtDNA test.

Anyone who has been researching their family tree should try the autosomal DNA test.  Your tree will grow.

DNA Link for Newark and Chicago Branches

Someone with a Lutter ancestor has appeared in my DNA matches (autosomal) from AncestryDNA.  This has never happened before.  (Even the Y-DNA test produced no matches.)

Possible relative in AncestryDNA matches for Jody
(Name and picture blocked by Jody)

The family tree for this DNA cousin had one Lutter ancestor, Emma Lutter, from Illinois.

Emma Lutter (1892-1948) in the family tree of Jody's DNA cousin

I know this Emma Lutter.  She lived in Chicago and married Frank Scaar.  I have Emma's bank book from the year 1907.

Emma Lutter's bank book.  Gustav Schwabe was her guardian.

The bank book for Emma and her siblings, Adolph and Gertrude, were not handed down to me as family heirlooms.  They were purchased on eBay.

The bank books caught my attention because I researched a man named Alex Lutter from Chicago.  Emma, Adolph, and Gertrude were his children.  I was looking for Alex Lutter because he witnessed the marriage in 1888 in Newark, New Jersey, of my great great grandparents, Hermann Lutter (1860-1924) and Clara R Uhl (1864-1955).  

Signature of witness Alex Lutter

Aside from this marriage record and the newspaper announcement of the nuptials, I found no traces of Alex Lutter in Newark, New Jersey.  Instead, I found a man named Alexander Lutter in Chicago.  In 1890 he naturalized, registered to vote, and married Ottilie Dahlke.

cookcountyvitalrecord.uscertificates.com
(fee-based site)

Alexander Lutter died in Chicago on December 23, 1897, age 33.  (Born about 1864.)  His wife died May 23, 1904, leaving the three children orphaned.

Alexander Adolph Lutter (1895-1969), the son of Alex and Ottilie, married Anna Kabitzke.  Anna's family contacted me.  The couple had no known children.  (Her relative was the first to write a guest blog post.)

Alexander the son filed for a passport.  Below is his photo from the application.  Alexander wrote that his father came to the United States from Germany in 1885 and resided in Chicago, Illinois for "12 years, uninterruptedly," from 1885-1897.

Alexander Adolph Lutter (1895-1969)
Is there a family resemblance?

If the father Alexander lived in Chicago from his arrival in 1885 until his death in 1897, this contradicts the voter registration from the year 1890, where Alexander is stated to have lived in Illinois for only one year.  Then I considered the source and decided that this information was not too reliable:  the informant was not even born until 1895 and was two years old when his father passed, so he likely has no personal knowledge of his father and no older relative to ask.

So how is Alexander Lutter (1864-1897) of Chicago related to my Lutter line of Newark?  We don't know- yet.  Let's hope that the submitter of the DNA contacts me and provides more clues.

With a birth date in 1864, Alexander Lutter could be the brother of my great great grandfather, Hermann Lutter, who was born about 1860.  In his will in 1924, Hermann mentioned two siblings, both deceased:  his brother Otto (1845-1909), formerly of Harrison, Hudson County, New Jersey; and his sister, Ottillia, formerly of Neuhaus, Thueringen, Germany.  If Alexander is another sibling, why was there no mention of Alexander or his children, three of whom were alive in 1924?

Surviving Genes of Walling and Dey

I have figured out the most recent common ancestors of a DNA cousin.

A new Shared Ancestor Hint appeared for me at AncestryDNA.  This system compares the family trees of my DNA cousins, looking for the same ancestors from my family tree.

The shared ancestor suggestion was William Walling (1770-1824) of Monmouth County, New Jersey.  He is one of my 5X great grandfathers.  (His transcribed will can be found here.)  I don't know why his wife, Rebecca Dey (born about 1770), my 5X great grandmother, did not appear also.

Based on our family trees, this DNA cousin is a fifth cousin of my father.  AncestryDNA reports that I share some DNA with this fifth cousin.

Here is the step where a lot of people stumble:  this scenario does not prove that William Walling and Rebecca Dey are our biological ancestors; nor does it prove that the DNA we share is from William Walling and Rebecca Dey.

We need to see how much DNA and which segments of DNA are shared with this DNA cousin so that we can triangulate the match.  You cannot triangulate with AncestryDNA.

To triangulate, we need to look at other DNA cousins who match on these same shared segments of DNA and determine if they too trace ancestry back to William Walling and Rebecca Dey, or one of the ancestors of either of William Walling or Rebecca Dey.  With AncestryDNA, you have no information about where any of your other DNA cousins match you, so you cannot know if they match on the same segments as any other DNA cousins.

I remembered the name of this DNA cousin from 23andMe, where we had corresponded two years earlier.  We had discussed the idea that the common ancestors were not too far back in time and were likely in Monmouth County, New Jersey because this DNA cousin matched my father, his siblings, and their father's cousin from the Winterton/Dunn line in Monmouth County.  If this segment can be further identified as a Walling/Dey segment, then we know that the segment came to my father via William Winterton's mother, Sophia Walling.

We can see the shared DNA at 23andMe.

Note that one of the matches, a sibling of my father, shares a very tiny segment of only 5 cM.  Most people would disregard such a small amount, but as you can see in the graph, this small segment is in line with the larger segments shared by other close relations.

Prior exchanges with this DNA cousin at 23andMe did not lead to William Walling and Rebecca Dey because his family tree had not yet grown that many generations back.  Two years later, his family tree at Ancestry was more robust and prompted the discovery.

The next step is to look at the other DNA cousins who share this segment.  I maintain Excel spreadsheets for my closest relatives.

This snippet is from the spreadsheet of my first cousin, twice removed.  I list his DNA cousins by chromosome of where they share identical DNA.  The Walling/Dey cousin shares an identical segment on chromosome 7, starting around 91,000,000 and ending around 127,000,000.

I checked all of these smaller matches against the Walling/Dey cousin to see if they match him as well as my first cousin.  All of them do.  If this shared DNA is from the Walling/Dey couple, then all of these DNA cousins will also descend from Walling/Dey or from one of the ancestors of either William Walling or Rebecca Dey.

This is where you need to depend on others for information.  Some of these DNA cousins offer family tree information; most do not.  I reviewed their surname and location lists and family trees, if they had any, and saw nothing in common with Walling/Dey.  I sent them all a note, describing what we need to look for.

So far, nobody has replied.  Without finding a connection to Walling/Dey in these other matches, we cannot triangulate to declare that this segment of DNA is from Walling/Dey.