It’s Another First Cousin

Roberts DNA
By Don Taylor

Photo of Don Taylor with cat Nasi.One of the benefits of using Ancestry DNA for Genetic Testing is their vast database.  Because there are so many people in their system, you are much more likely to have a DNA match. Sure enough, it happened again. This time, a previously unknown person, Debra contacted me via Ancestry Messages with the simple message, “My DNA results says that you are my 1st cousin.”

Oh my, here we go again.

I clicked on View the Match, then clicked on the little “Info icon” to see how much DNA we shared. Debra and I share 621 centimorgans across 25 segments. According to the chart I use, that amount of shared DNA put us in an overlapping range of first cousin and first cousin one removed. I then clicked on “Shared Matches” and saw that she also matched with my Roberts half-siblings. Because I can view my half-sister’s matches, I looked at her results and saw that she and Debra share 893 centimorgans of DNA across 37 segments. Solidly in the first cousin range. For sure, Debra is a first cousin and now I knew that we share a common grandparent on my paternal side.

My grandparents, Bert Allen Roberts and Essie Pansy Barnes, had five children. The amount of DNA shared was not enough for Debra to be my half-sibling, so that ruled out my biological father, Hugh Eugene Roberts, from being involved. In subsequent messages, she indicated she knew who her mother was, so that eliminated Pansy and Helen, leaving only two potential sources for her to be a first cousin – Uncle Bert and Uncle John. Between the two, Uncle Bert was, by far, the likely candidate.

Photo of Bert Allen Roberts, Jr with two (unknown) women.
Bert Allen Roberts, Jr. and two unknown women, c. 1947.

Then, Debra let us know that her sister told her that her father’s name was Bert, but never knew his last name. Debra also sent a photo of Bert, her supposed father, from the late 1940s. My half-brother Tom knew Bert and was able to identify Uncle Bert from the picture.  Mystery solved!

So, welcome cousin Debra Edwards to the growing Roberts clan. I am so pleased you were able to identify who your father is after so many years.

So far, DNA test results have led to my learning about:

Note: I wish Family Tree Maker had a better way to indicate offspring producing relationships.  Creating a “spouse” and then set the relationship set to “Friend” or set to “Other” is cumbersome at best but doesn’t describe the relationship. Sigh….



Glennis Paternity Project / Roy L Huber?

Glennis Paternity Project

By Don Taylor

Image of DNALast fall I began a project to look more closely at the possibilities of figuring out who my half-sister Glennis’ biological father might be. Her closest match was a predicted second cousin, Mike. Mike had a nice, full tree.  If Mike and Glennis were 2nd cousins, they would share great grandparents.  I took each of Mike’s eight great-grandparents and followed their descendants to see if any of them might have been at the right place at the right time to have fathered Glennis.

See the following articles:

Only the Hemsworth-Morgan Branch yielded any potential candidates.


This month I looked at the closest paternal matches that Glennis has on GEDMatch. The closest match to Glennis was J.C.  I contacted the person managing J.C.’s kit, and he provided information for me to see his tree.

I knew I was getting close when I saw that Mike’s ancestors and J.C.’s ancestors both included the surnames Odell and Morgan and both in West Virginia. I researched J.C.’s ancestors back one more generation and found that J.C.’s 3rd great-grandparents were the same as Mikes. Four of the 3rd greats were the same. What occurred is that a brother and a sister married a sister and a brother.  That makes their children double first cousins. Double first cousins should share about double the genetic material in common than regular first-cousins. As I understand things, that should push the expected common ancestor back one generation.

So, the good news is that following J.C.’s line back has convinced me that Jacob M and Elizabeth (Smith) Morgan and Joshua and Susannah (Davis) Odell are common ancestors to both Mike and J.C. Likewise, I’m confident that Glennis is descended from them as well. The bad news is that because of the double first cousin status, I am not sure which of Nathan and Belinda (Odell) Morgan’s eleven children’s descendants are the likely ancestors. It could be any of them.

I had followed one of Nathan & Belinda’s children, Mary D, and discovered two potential candidates to be Glennis’ biological father. Now, absent any new DNA Matches, I will need to follow the other ten children’s descendants looking for potential candidates.

Roy L. Huber (1932-1999)

(Potential Candidate (from the Hemsworth-Morgan Branch) analysis.)

Roy L. Huber was born on 31 August 1932 in Marietta, Washington County, Ohio (although his death record indicates he was born in Michigan).[i] He was the youngest of two (known) children of John Clifford Huber and Naomi M. Stewart. (The great-grandson of Mary D. Morgan and the 2nd great-grandson of Nathan and Belinda (Odell) Morgan.)

The 1940 Census indicates that the Huber family lived in Dearborn, Wayne County, Michigan in 1930 but had moved out to Milford township by 1940, about 35 miles away.[ii]

Emblem of the Michigan Air National GuardHis whereabouts in early 1953 are unknown, but I presume he was in Michigan. (When Glennis was conceived.) In 1956 and 1957, 2nd Lt. Roy L. Huber was serving with the Michigan Air National Guard. He was still in Michigan in 1990 when he appears to have invented an “auxiliary vehicle deck” for pickup trucks.[iii], [iv], [v]

He appears to have started “snowbirding” to Florida in the 1990s as he seems to have had addresses in both locations during that time.  He died in North Palm Beach, Florida, on 6 November 1999.[vi], [vii]

I have not found an obituary for him.  Nor have I found any evidence that he ever married or had children.  He is still a possible candidate.

If you are a close relative of Roy L. Huber, I would love to hear from you through the contact form below.


  1. Research the other ten children of Nathan and Belinda (Odell) Morgan and determine if any are potential biological fathers for Glennis.
Children: Sex Birth Death
Elizabeth Morgan F Bet. 1852-1854
West Virginia
Eli H Morgan M Abt. 1855
West Virginia
Mary D Morgan

(Completed already)

F 14 Jan 1858
Pleasants, WV
25 Apr 1939
Williamstown, WV
Isiah S Morgan M Bet. 1859-1861
West Virginia
Francis Marion Morgan M 1840 1922
Sanford Morgan M 1843
Jacob Ellison Morgan M 1850
John Wesley Morgan M
Asceneth Morgan F
Jane Morgan F
Sarah Morgan F

2. Research Roy L Huber’s brother and see if he remains a candidate.


[i] United States Public Records, 1970-2009, Family Search, Roy L Huber – 31 Aug 1932. Accessed 4 July 2016.
[ii] 1940 Census, Family Search, John Huber – Milford Township, Oakland, Michigan – Line 18.
[iii] 1998-06- – Page 17 – Letters – Airplane Noise – Roy L Huber. See File: 1998-06-26 – The Palm Beach Post (West Palm Beach, Florida) · Page 17 Letters – Airplane Noise – Roy L Huber.pdf., Palm Beach Post, The, West Palm Beach, Florida (
[iv] 1957-05-07 – Page 1 & 2 – Roy L Huber [Last Sentence]. 2nd Lt. Roy L. Huber represented the Michigan Air National Guard at an exhibition of aerial firepower and tactical maneuvers at Eglin Air Force Base, FL ., Battle Creek Enquirer, Battle Creek, Michigan.
[v] “United States Patent,” The United States Patent and Trademark Office, US Patent 4962709 – Roy L. Huber – Auxiliary Vehicle Deck.
[vi] United States Public Records, 1970-2009, Family Search, Roy L Huber – 31 Aug 1932. Accessed 4 July 2016.
[vii] United States Social Security Death Index, Family Search, Roy L Huber (1932-1999).

OMG – Another Half-Sibling

Half-Siblings provide the proof

Brown, DNA
By Don Taylor

Photo of Don Taylor with cat Nasi.Thanks to autosomal DNA testing, I’ve learned who my biological father is. I have discovered and met some of my “new” half-siblings on my biological father’s side. I have also discovered that my wife has a previously unknown half-sister. Now, due to DNA testing, I’ve found that my mother has a previously unknown half-sister.

It began with an email from (I’ll call her) HC, who indicated that Ancestry DNA was saying that she and I were first or second cousins. The Ancestry match reported that she and I share 460cM of material.  A look at our trees showed no surnames in common. Ancestry allows you to view a match and see who also shares that match.  My half-sister, Glennis, was also a match and shares, even more, DNA (522 centimorgans) than I share with HC.  That proves that the match was on my maternal side as Glennis and I share a common mother.

Screen shot showing "HC" and author share 460 centimorgans of DNA.
HC & I share 460cM

Through an exchange of messages, I learned that HC’s mother was adopted, was born in May of 1938 in Texas, however, her mother was conceived in Minnesota. That narrowed things considerably.  My mom’s Montran/Barber line pretty much was from Michigan; my mom’s Brown/Manning line was from Minnesota. So, it was very likely that the match came from my mother’s father’s side of the family.  Luckily, my mother has a half-sister.  The bad news is that neither my mother or her half-sister, Barbara, tested with Ancestry.

No problem, GEDMatch to the rescue. Although both tested with another service, I had previously exported their data from the other system and imported the data into GEDMatch. If HC was a match with my mother and aunt Barbara, then the common ancestor had to be on their common father’s side. If the match was only with my mother and not my aunt Barbara, then the common ancestor had to be on her Montran side. I know very little about Montran line, so anything could be possible.

HC uploaded her data to GEDMatch and the results were amazing.  She shares over 1000 centimorgans of DNA with BOTH my mother and my aunt Barbara – Proof that the common line is on the Brown side. I like to use The DNA Geek’s chart to quickly see the potential relationships between individuals at a particular centimorgans level. The chart shows that 1000 cM is solidly in the range of Group C relatives. Relationships for Group C include First Cousin, Half Aunt-Uncle/Niece-Nephew, Great-Grand Parent/Child and Great Aunt-Uncle/Niece-Nephew.

Now that I know that the match is on the Brown line I can speculate.

Grandpa Dick
  • If Grandpa Dick is the father of HC’s mother, then HC would be the half-niece of my mother and Aunt Barbara.  That fits the amount of DNA Perfectly.
  • If one of Grandpa Dick’s brothers were the father of HC’s mother then, HC and my mother would be first cousins once removed and I would expect a DNA match of between 215 and 650.
  • Dick’s father died in 1928, so he can’t possibly be the father of HC’s mother, so that scenario isn’t possible.

Finally, I questioned was there is a locational opportunity for Grandpa Dick to be the father. HC’s mother was conceived while her mother was in Deerwood, MN about August of 1938. In 1937, my Grandpa Dick was living in Brainerd, Minnesota, about 18 miles from Deerwood.

I think that is enough to prove the relationship. However, I always like to go the extra mile if possible and prove it beyond the shadow of a doubt. HC’s mother is still alive and recently had her DNA tested. When the results come back, we can confirm this relationship. I expect that the autosomal DNA match with my mother and with Aunt Barbara will be in the 2000cM range – solidly in the half-sibling range.

Additional proof will come through a comparison of the X chromosome. Females have two X-chromosomes (males have an X and a Y).  One of the X chromosomes is from the mother and is recombinant, that is to say, it is a mix of the mother’s X.  The other X chromosome is a replica of the father’s X and is passed on without change.  If HC’s mother and my mother are half-siblings, I would expect to see their X-Chromosome to have a solid match like my mother and her half-sister Barbara have.

Screen Shot - X Chromosome Match of 2 half sisters
X Chromosome match of my mom & Aunt Barbara.


My mother and my Aunt Barbara have a here-to-for completely unknown half-sister. Amazing. I always heard that Grandpa Dick “liked the ladies.”  I guess he did. I now know of four daughters that he fathered, my mom, Aunt Barbara, Aunt Mary Lou, and newly found Aunt Phyllis. I wonder if there are more….


  • I do not typically use the full name of living individuals.
  • Of course, if any DNA specialists see anything incorrect with my reasoning above, please let me know via the contact form below.

Keep Trees Wide, Not Deep – Example: Mannin/Barnett

Brown-Montran Research
DNA Research


During the last meeting of the Maine Genealogical DNA Interest Group, someone asked if it is better to have a tree that is deep or a tree that is wide. I mentioned that, for autosomal DNA test matches, a wide tree is best.  The sheer number of potential 5th and 6th cousins is daunting. But, more importantly, the likelihood of your sharing DNA with a 4th cousin is only 69% and the likelihood of sharing DNA with a 5th cousin is only 30%.[i] Consequently, knowing your 10th great grandparents is of little use in matching DNA cousins.  (Consequently, knowing your 10th great grandparents is of little use in matching DNA cousins. There are two exceptions to this, Y-DNA tree (paternal only) is useful for connecting trees on a Y-DNA match.  Also, X-DNA can provide a similar usefulness.)

23 & Me Shared Matches
23 & Me: Shared Matches

The importance of having a wide tree was exemplified recently.  I was contacted through 23 and Me by a, potentially, 2nd to 4th cousin (I’ll call B.J.) I took a look at the match using 23 & Me‘s new She and my aunt Barbara shared 88cM across five segments. My mother shared 50cM across two segments; interestingly enough, I also shared 50cM across two segments. Looking at what segments all four of us share is an excellent example of how sticky DNA segments are.  All three of us shared the same sticky chunk of DNA.

Screen Shot - Chromosome 3 comparison
Screen Shot – 23 & Me – Chromosome 3 comparison showing sticky clump shared among all of us.




We exchanged basic tree information, she mentioned her ancestors were a Mannin and a Barnett. When she said that, I knew we were related and I was pretty sure I knew exactly how.  Nancy Ann Mannin married Jessie Monroe Barnett about 1867 in Kentucky. They later moved to Minnesota and settled May Township in Cass County, Minnesota.

A couple more email exchanges and I learned that B.J. and my Aunt Barbara were third cousins their common ancestor was Enoch Mannin. Enoch was one of those pivotal people in my genealogical research and I knew a lot about him and his descendants. I even had B.J.’s mother (but not her father nor her) in my family tree records.

Thanks to 23 and Me for providing the tools to connect with another cousin.

———-  DISCLAIMER  ———-

I have tested my mother, my aunt, and myself with 23 and Me – Have you?


[i] Internet: DNA Land – “Face it: DNA cannot find all your relatives”

Peterson Paternal Project – Biddle-Hall Branch

Part Four

Sometimes it just gets easier, but the results are difficult to believe, so you keep making more work for yourself.  That is what happened when I investigated the fourth line of my Peterson Paternal Project.[i]

Samuel Biddle and Mary Margaret Hall – Descendants

I needed to follow the descendants of Samuel and Mary Margaret (Hall) Biddle and see where any male descendants were likely to have been in 1953 when my half-sister Glennis was conceived. I kept looking, and looking, and looking for more children and grandchildren for Samuel and Mary, but just couldn’t find any.  If I had looked at the family trees of other people and believed them when they showed only one child and four grandchildren, I would have saved a lot of time and effort.

Descendants of Samuel Biddle and Mary Margaret Hall

Samuel and Mary had four children, however, three of them had no issue. Only their youngest child, Nicholas Boulden Biddle, had children – four of them.  Nicholas’ daughter, Mary Jean, was too young to have had male children old enough to be the biological father of Glennis.  That left Nicholas’ three boys, Richard Wyllys, Brent N, and Samuel Lee to have the potential to be Glennis’ biological father.

First, if any of the three were the biological father, then the relationship would be first cousins rather than second cousins.  The DNA cMs indicated 2nd cousins, not 1st so, this is a very unlikely connection.

Second, all three of them married before 1950.

Third, all three of them were in West Virginia in 1949.

Finally, all three appear to have stayed in West Virginia into the 1950s and none of them show any indications of having left West Virginia in 1953.

Consequently, I don’t believe any of them is a candidate to be Glennis’ biological father.


[i] My half-sister Glennis has long wondered who her biological father is. Ancestry DNA provided a clue; she shares 201 centimorgans of DNA across 8 segments with someone.  The shared CM suggests they are second cousins, meaning that she and the other person likely share a great grandparent. The logic is that if I can take all eight of his great-grandparents and follow their descendants, possibly one of them was in the right place at the right time to be Glennis’ father. If so, I will have a very likely candidate to my Glennis’ biological father.