Tuesday, May 31, 2011

FTDNA's Family Finder: My New Illumina Chip Results and Investigating a "Close" Match

I finally received my updated Family Finder results with the new Illumina chip from FTDNA. There was a problem with my sample, so it took much longer than expected. I have been excitedly investigating my new and improved match list.

With FTDNA's original Affymetrix chip I had 54 total Family Finder matches. I now have 59 total matches with the new Illumina chip (including some new customers). Between the two chips, there are only 22 matches in common and 32 matches from my old list are gone completely. On the Affy chip I had three matches under the "Close to Immediate" match filter. I now have four using that filter on the Illumina chip. One is the same. One is a brand new participant. One was previously listed as 4th to distant and one was previously listed as 5th to distant. Of the old Affy "Close and Immediate" matches, one is now listed as a 4th to distant and one is completely gone from my match list.

I was surprised and happy to see the brand new "close" match.  This is my first 3rd cousin prediction on Family Finder (range 2nd - 4th). We have 2 blocks of DNA in common. The one on Chromosome 5 is 17.67 cMs and on Chromosome 18 we share 5.63 cMs. My match agreed to upload her raw data file to GEDmatch, so I could compare her to my mother's 23andMe file. This allowed me to determine that they match on longer stretches of the same spots: Chr 5 = 21.9 cM and Chr 18= 9.4 cMs. I share ~ .315% of DNA with her and my mother shares ~.42%. Since a 3rd cousin will share on average .781% and a 4th cousin .195%, this one appears to fall somewhere around a 3rd cousin once removed.  (See expected percentages here.)

Yellow bars are where I match my new predicted 3rd cousin

We confirmed that we do not have any second or third great grandparents in common, so the match must be further back than predicted. We haven't been able to figure out our exact connection, but we have narrowed down our match to a very specific area and branch of our trees. Her fifth great grandmother, "Catherine" married Johannes (John) Long. Some believe Catherine was a Kinard, but she has long been rumored to have been a Roderick. If so, this would mean that she is related to my third great grandmother Susannah Roderick, since all in the area with that surname were descended from the same immigrant Rothrock family. Catherine is the right age to be the sister of Susannah's father Daniel Roderick. If this is correct, then their parents Johann George Rothrock and Elizabeth Roemig would be my match's sixth great grandparents and my fifth great grandparents.

My match also theorized that Susannah's mother Elizabeth Long is the sister of Johannes Long who married Catherine. If indeed Johannes Long and Elizabeth Long are siblings, then their parents (unknown) are her 7th great grandparents and my 5th. This would make us double 6th cousins once and twice removed.

This double cousinship could account for the surprising amount of DNA that we have in common. Presently, this is just a theory but, certainly, our close match lends credence to Catherine, indeed, being a Roderick/Rothrock. These Pennsylvania Dutch families all lived in Bucks and Berks Counties, Pennsylvania, intermarried and traveled together to (Ross County) Ohio and Illinois. They may have even known each other back in Germany. Although, I am not totally convinced that we have determined our exact connection, I am confident that we have pinpointed the correct family branches and geography. 

There is also a possibility that we have a connection through Susannah Roderick's husband, my brickwall Asa Travis. Since these families intermarried frequently, it is conceivable that his parents were also related to one of the Pennsylvania Dutch families traveling together. There is little doubt that multiple common ancestors and, likely, cousin intermarriages account for our "close" match. I will be keeping watch for overlapping matches that may help us to narrow down the surnames involved in these connections.
[Disclosure - My company StudioINTV has an existing production agreement with FTDNA that has no bearing on the opinions I express. I also receive a small commission from FTDNA on non-sale orders through my affiliate link, which I use to fund DNA tests. I receive no other compensation in relation to any of the companies or products referenced in my blog.]

Monday, May 30, 2011

Confirmed another 23andMe cousin today!

Yesterday one of my friends/cousins at 23andMe mentioned a predicted cousin of his who has ancestors from the Gloucestershire area of England. Since my Allen, Rudge, Trotman, Chitts, Hill, Johnson, Fowler, Bailey, Jones and Pullin families are from that area, I checked out his profile. Of the seventeen surnames on his profile, we shared four (Pullin, Allen, Cornell and Gillet), so I sent him a share request.  Sure enough, he matched both of my father's siblings on Chromosome 13 for a 6 cM stretch.

Click on the chart to enlarge and get a closer look

With confirmation of a match, I immediately focused on the Pullins since according to his profile, he had traced them back to Somerset, UK, which isn't too far from where my Rachel Pullin (b. abt. 1781) married William Jones in 1801. Because I was concentrating on them, I completely ignored the fact that he also had mentioned the area of Waterloo, Ontario (Canada).

After looking at "My Tangled Vine" family tree on Ancestry.com, he informed me that Sylvanus Cornell (b. abt. 1752) who emigrated from New York to Waterloo, Ontario in about 1800, is his 4th great grandfather.  Sylvanus and his wife Joannah are also my dad's 4th great grandparents. This makes my dad's siblings and this match 5th cousins, fitting with 23andMe's prediction that they are 7th cousins with a range of 4th - 10th. [Update - Sylvanus is also his 5th great grandfather, making them 5th cousins once removed as well.]

It is difficult to be sure if our Cornell ancestors are responsible for this match since we very well may also have Pullin and Allen ancestors in common. Also, we do not know Joannah's maiden name and the shared DNA could be from her. So, as I have said before, to be 100 percent confident that we have found the correct connection responsible for the shared DNA, we will have to wait and see if we get any other matches on this spot who also have Cornell, Allen or Pullin ancestors.

One especially unique thing about today's match is that we have so many possible connections. He shares Gillet with my maternal side, he shares Cornell with my paternal side and we might even share Pullin and Allen ancestors from the area around Gloucestershire, England. (His 3rd great grandfather is Isaac Allen and so is my dad's. Not sure if it is the same one yet.)  To make it even more interesting, I have previously corresponded with his nephew (the curator of the Sheffield Museum) about family history, he belongs to David Faux's Young Family Project that I have mentioned in this blog here and he has lived for forty years in Sheffield, where my fourth great grandfather Henry Moore lived and is buried, even attending the church that my Moores helped the Cornells to found in Beverley, Ontario two hundred years ago. When you find a connection like this, it makes all the hard work worthwhile!

Sunday, May 29, 2011

Autosomal DNA Matching and the Importance of Testing Multiple Family Members

I was contacted by a new cousin at 23andMe. She noticed a matching surname on our lists and requested to share genomes. Upon doing so, we discovered that we do indeed have a match. She matches me on 6.5 cMs of Chromosome 14 and Relative Finder predicts that we are seventh cousins. We quickly determined from our family trees that we are 9th cousins once removed, which slides in right at the end of the predicted range of 4th - 10th cousins. Our common ancestral couple is John Spofford (b.1678) and Dorcas Hopkinson (b.1676) from Massachusetts. Done. Great! Right?

Not so fast. Upon further investigation, I noticed something very interesting. The common ancestral couple is on my father's side, but my mother shares this match with me!  This "Spofford" Cousin matches not only me- but my mother, my sister and my two nieces on Chromosome 14. Therefore, the match on Chromosome 14 must come from my mother's ancestry, not my father's. So, is this once promising match a total failure? Take a look at Chromosome 5. That little dark blue smudge is a match that my sister shares with this cousin, but my mother and myself do not. Now, that one could be from my dad's side!

For a closer look, please click on the charts to enlarge

Actually, this Spofford Cousin matches both of my sisters on Chromosome 5. Dark blue and light blue smudges show up on Chromosome 5 on the chart below, signifying that my sisters share this 6 cM match.

Since I do not have my father's data at 23andMe, but I have tested his two siblings there, I checked to see if either one matches this cousin. Neither one does. This makes it impossible at this time to absolutely determine if this stretch of matching DNA on Chromosome 5 was inherited from the Spoffords. If, in the future, other matches show up that are descended from the Spofford or Hopkinson Families that will clarify the situation. Since I do have my father's data over at FTDNA's Family Finder, another possibility is to compare his raw data file with his Spofford Cousin at an independent project that compares data files from both companies, like GEDmatch (if she is willing to submit).

Because of the matching DNA on Chromosome 14, I examined my Spofford Cousin's family tree a bit more to see if I could determine from where that match might originate. I couldn't help but notice the surname Shellenbarger figuring prominently on her tree. I have written in the past about my mother's mysterious Stolebarger great great grandparents here and here and my theory that their often misspelled surname with its many variants could have evolved from a similar name. Sure enough, both the Shellenbargers and the Stolebargers were German immigrants living in Pennsylvania in the early 1800s and members of both families ended up in Iowa by the late 1800s. I could not find a solid connection, but this match will cause me to be more keenly aware of my mother's links to these Pennsylvania Dutch families, including our Roderick, Long and Roemig ancestors. This may even lead me to finally start that Stolebarger Y-DNA Project. In fact, I still have an unused kit sitting here waiting for a Stolebarger, Stolabarger or Stoalabarger male. Any takers?

Saturday, May 28, 2011

Known Relative Studies with 23andMe: Second Cousins

I recently received the results for my mother's second cousin. Our common ancestral couple is Matti Yrjanasson Syrjala Wiita (1852-1931) and Justiina Johanssdotter Wiita (1843-1907), both born in Ylistaro, Finland. He descends through their daughter Josefina and my mother and I descend through their son Matti.

I am fortunate to have a number of relatives with whom to compare this cousin. The expected percentage of DNA in common with a known second cousin is 3.125%. He and my mother share 3.27%, while he shares about 3.1% with her sister and 2.77% with their first cousin V.  It is interesting to see in the chart below that each of them inherited mostly unique segments with only a few segments in common. The chart compares my mother in dark blue, my maternal aunt in light green and their first cousin V in light blue to their second cousin (KnownC). All three women have the same familial relationship to him, but due to the random inheritance of autosomal DNA, they each share unique patterns and amounts of DNA with him. Notably, all four only share at one small spot on Chromosome 11.

Click on the chart to enlarge and get a closer look

I am also able to compare our known cousin to descendants of the women above to demonstrate what occurs when the DNA is passed to successive generations. Second cousins once removed would be expected to share half as much DNA as a second cousin (1.563%) and a second cousin twice removed would share half of that (.781%), the same percent as a third cousin. My sisters and I are all second cousins once removed from our known cousin. I share 1.08% and  my sisters share 1.84% and 1.38%. For second cousins twice removed, the actual percentages found in this study are .38%, .73%, .84% and .57%.  Below are four charts comparing our cousin to various family members, demonstrating how the DNA inheritance changes as it is passed down. All of the charts are laid out in the same way - comparing one of the second cousins and then two successive generations. All charts can be enlarged by clicking on them.

Note: On Chromosome #8 there is an unaccounted for small match. This match does not show in any other family member and may be a psuedo-segment
Note: On Chromosome #22 there is an unaccounted for match, This may be from inherited from the paternal side.

**Update - In the comments below, Dr. Ann Turner makes an important point about these four charts. She writes, "In addition to showing how shared segments vary in size and position, the multi-generation diagrams also show how the segments are inherited in an all-or-nothing fashion much of the time, especially as the segment size grows shorter. The longer segments naturally have a higher probability of being split by recombination."

Monday, May 16, 2011

Postage increase affects delivery of FTDNA kits

Apparently, there has been a large backlog of undelivered mail intended for FTDNA sitting at their local post office for the last couple of weeks. This was due to the fact that postage rates increased on April 17, 2011 for all types of mail except first class single ounce letters, thus changing the amount needed to return the DNA kits (previously $1.22). If you are wondering why you have not received notification that FTDNA has received your kit, yours may be one of those awaiting postage due. I am told that FTDNA is correcting the problem and having that mail delivered tomorrow as well as updating the recommended postage on the kit enclosure to reflect this increase. Meanwhile, if you have an unused kit at home, please be sure to note this change. The cost for each additional ounce after the first has increased from 17 cents to 20 cents. The post office recommended one customer post $1.71 for the return of the kit. I would suggest having yours weighed just to be sure.

**Update - FTDNA has confirmed that the correct postage to return their kits is $1.71.