Friday, August 28, 2009

Godwin-Boykin DNA puzzle

Last week I got an email from someone who matched my grandfather's DNA on Ancestry.com. Her husband's surname is Boykin and my grandfather's surname is Godwin. My grandfather's Godwin DNA matches at least 8 other Godwin test participants, so I'm fairly certain he's really a Godwin. However, we are not so sure that this Boykin match is, genetically speaking, really a Boykin. His great-great grandmother, Barbara Boykin, appears to have been a true Boykin, however, as we know, the y-chromosome DNA test only applies to the direct male lines.

Barbara Boykin was found on the 1850 census in Johnston Co., NC as an 18 year old girl living in the household of Jesse and Elizabeth Boykin. There is nothing unusual about this census report, therefore I would assign a preliminary relationship of Jesse being Barbara Boykin's father. Also, I would assume that Boykin was Barbara's maiden name.

So what of her son, Perry Boykin, born about 1856 in North Carolina? There is no mention of a father. Barbara and her young son were living with the Barry Johnson family on the 1860 Neuse River, Johnston Co., NC census report. In 1870, Perry Boykin was living with another Johnson Family, this time in Old Fields, Wilson Co., NC. His mother Barbara was not living with them. By 1880, Perry Boykin had married and was enumerated with his wife, Wealthy Jane and daughter Sarah in Old Fields, Wilson Co., NC. His mother Barbara Boykin, a single woman, was also living in the house.

The Johnston County Court minutes should be reviewed for any "bastard bonds" that might have been issued to Barbara Boykin for the maintenance of her son (who did not have a father). Wynette Parks Haun has written several abstract books for the court minutes of Johnston County, but I'm not sure if any of them go up to the year 1856 when Barbara Boykin's son Robert Perry Boykin was born. If not, then the microfilm will have to be reviewed. I hope it has an index!

With regards to the DNA....

My grandfather tested with ftDNA's 37 marker test. This Boykin match tested through Ancestry.com with their 46 marker test. Of those 46, 32 markers intersected, however we were not a perfect match. We were a 31/32 marker match. Ancestry.com gave us a 95% probability that our common ancestor was about 600 years back, not really genealogically significant. However, there is a 50% probability that our common ancestor was only 200 years back.

So did his Boykin great great grandmother have a child with one of my Godwin ancestors? My cousin Lori did find one record between Thomas Godwin and Thomas Boykin in 1828 in Johnston County, NC, however it is unclear if this Thomas Boykin had any ties to Barbara Boykin or her supposed father, Jesse Boykin. She also found another older record dating back to Isle of Wight County, VA, where both of our families are from: A Frances Godwin married General Frances Marshall Boykin in 1805. This Godwin family is not believed to be related to my own however.

Thursday, August 6, 2009

The Genetic Distances between Markers and the Affect it has on MRCA

Family Tree DNA has a great article published on its website about deciding how many markers to test in order to determine the time to the most recent common ancestor.

This is a screenshot of the MRCA statistics table on their website below:



I was trying to determine by how much the number of generations to the common ancestor increased as the number of genetic distances (differences between markers) increased. Based on the number posted above, I predicted the following:
It looks as if you have only 1 marker difference, the number of generations to your most common ancestor increases by 1/2 of the number of generations you started with.

For example, if you have a 37 marker test, and you are a 100% match, then there is a 95% probability that you shared a common ancestor within the last 7 generations. If you match 36/37 (1 marker difference), then the number of generations increases by 1/2 of the original number: so you started with 7 generations, take 1/2 of that which is 3.5 and add it to the original number = 10.5. So the number of generations to your common ancestor just increased from 7 to 10.5! If there is a 2 marker difference you would add another 3.5 to the 10.5 = 14!! These numbers are reflected in the table above and the calculations seem to apply to all tests.

I believe the standard for number of years per generation is about 45 years. So 14 generations would be about 630 years. Surnames only started about 600 years ago (~ 1400's), so anything beyond 600 years probably has little genealogical value.
I referenced Genealem's Genetic Genealogy Blog post article, "Why test 67 markers?" to come up with the standard of about 46 years / generation.
Genealem wrote: "25 marker match gives you a 95% probability of having a common ancestor within the last 600 yrs"
Combining the probability for a 25 marker match of 600 years and that of 13 generations as listed in the table above, the calcuated value is about 46 years / generation.
I originally wrote this article for family members who were trying to understand the genealogical value of DNA results. I am in no way an expert on DNA testing, but have developed a few tips and tricks along the way to better understand the results myself. This article reflects my findings.