Wednesday, February 19, 2014

Sneak Peek at the New Gedmatch.com Chromosome Triangulation Tool

For those of you who use Gedmatch.com to view, share, and analyze your DNA results, you might have noticed that the "Check for matches on a specific chromosome segment" utility was removed from the site. I used this feature most of all to find all my matches who matched me on a particular chromosome, but found lately that this utility was too slow to use.

Those of us who loved this tool will be in heaven as it has been replaced with a new tool called "Segment Triangulation." This new utility acts similarly to the previous utility with the added function of showing you matching segments BETWEEN the matches on every chromosome. 

Here is a sneak peek of what it looks like. 


Gedmatch - Chromosome Segment Triangulation Tool

More information on how to use this tool is to come. 

Have you check out the new Segment Triangulation Tool? If so, what did you think? 




Saturday, February 15, 2014

The Wrong Peters Family


In this post, I will discuss how James and Martha Peters were eliminated as potential parents of my 2nd great-grandfather, William Edward Peters via DNA Testing.

William E Peters and Stanley Lark
William E Peters and Stanley Lark
Pocahontas, Arkansas, c 1940

My 2nd Great-Grandfather, William Edward Peters:

William Edward Peters was the father of my great-grandmother Nova Peters. According to his obituary and death certificate, he was born 10 March 1874 in Thomasville, Oregon Co., Missouri and died 3 February 1948 in Pocahontas, Randolph Co., Arkansas. Unfortunately, his parents’ names were not listed on his death certificate or in his obituary, no family bible has been found containing information on him or his family, and there is no known information about any siblings he might have had. 

William Peters is my brick wall.

According to the Research Report I have written about my William Peters, I was looking for an 1880 census report that had a William Peters on it who was born about 1874 in Missouri when I came across a potential family. I found the 1880 Oregon County, Missouri Census report of Martha Peters with a 7 year old son named William Peters who was born in Missouri. James W and Ella were also enumerated with Martha and William. [1]

1880 Oregon Co., MO Census for Martha Peters
1880 Oregon Co., MO Census for Martha Peters


The Family of James and Martha Peters:

I couldn't find anything on Martha, James W., or Ella Peters between 1880 and 1900, nor could I find marriage or death records for Martha and a possible Peters husband in MO. I thought that since Martha was from TN and William's parents were from either  IL, TN or MO, [2] that maybe Martha and Unknown Peters married in TN prior to moving to MO and maybe they passed through IL on their way from TN to MO? I could not find the family of Martha Peters anywhere in Missouri in 1870. 

I did find a family with Martha Peters, born about 1844 in TN on the Anderson Co., TN 1870 Census with husband James Peters and son Elijah Peters. I reached out to the descendants of this family through message boards and learned that James and Martha only had 2 or 3 children, none of whom were the William, James or Ella who were listed on the 1880 MO census report with Martha.

I did not get discouraged though, and tried to find ways to make my William Peters fit in with this James and Martha family. James' father was named Tobias Peters and he also had a grandson named Tobias Peters. This was of interest to me because my great-grandmother's nickname was "Tobe," and no one knows where that nickname came from. I thought this was a plausible explanation.

Also, there is a huge mystery around the family of James and Martha Peters: The story goes that Martha and James left TN and moved to a farm in Arkansas. Then one day James up and disappeared leaving Martha to tend the farm and raise the kids alone. She finally sold the farm and she and her kids moved in with her parents in Sebastian Co., Arkansas. [3] Supposedly the new owners of the farm found a body on the premises which everyone agreed must have belonged to the missing James. However, according to letters written by James' children, several years later his children received a letter from him attached to a substantial amount of cash, indicating he was still alive at that time.

Putting aside the mystery surrounding James' disappearance, and looking at Martha Peters, his widow or spouse, I found that she was also missing. Although her descendants claim that she and her children moved in with her parents in Sebastian Co., AR, only her children were enumerated on the 1880 census report with her parents. This is why I wondered whether the Martha Peters I found on the 1880 Oregon Co., MO census report could have been the same Martha Peters. It seemed feasible to me that maybe she had remarried and had 3 additional children whom she was raising on her own while her parents looked after her older 2 children in Arkansas. The descendants of Martha and James that I corresponded with did not share my suspicions. One descendant did have a letter that mentioned Martha had remarried to a man named Williams, but there was no mention of any additional children.


DNA Doesn't Lie:

I finally put my suspicions aside about my William Peters being connected to the family of James and Martha Peters when I was contacted by another descendant of this family. She had read about my William on my blog and decided to write to me and introduce herself. 

She mentioned that she and her cousin had done the autosomal DNA testing and offered to compare our results to see if we were a match. Me, my mother, and Grandfather tested through ftDNA.
She tested with 23AndMe, so she had to upload her test results to Gedmatch.com for us to compare our DNA results since we tested with different companies. We also compared my family’s DNA results to her cousin, the granddaughter of Elijah Peters and great-granddaughter of James and Martha. Their grandmothers were double-cousins (mothers were sisters, fathers were brothers).

Although her cousin tested through ftDNA, she did not show up as a match to my family. I still did a comparision with Gedmatch.com just in case there was a small match that was not big enough to reach the threshold to be considered a match. [4]

In doing the Gedmatch.com comparison, we ran a one-to-one comparison between the first descendant's kit and mine, her kit and my mother, her kit and my grandfather. Then we did the same for her cousin's kit. We set the minimum cM value equal to 1. 


Here are the results: 


  1. Descendant vs Me - 0
  2. Descendant vs my Mother - 0
  3. Descendant vs my Grandfather - 0
  4. Descendant's cousin vs me - 4.3cM
  5. Descendant's cousin vs Mother - 4.9cM
  6. Descendant's cousin vs my Grandfather - 4.7cM
  7. X chromosome comparison between Descendant's cousin and my Grandfather - 0
  8. X chromosome comparison between Descendant and my Grandfather - 0


The amount of DNA that the descendant's cousin shares with my family is less than 5cM which is not enough to be significant and is probably just noise, therefore there is no connection between our two families. My family does not share any DNA in common with the descendant either, further supporting the conclusion that our two families are not connected. 

Conclusion: 

This is just one example of how DNA can either support or refute a hypothesis. In this case, our hypothesis was that my William Peters is connected to the family of James and Martha Peters. We compared the DNA results of my family  to those of James Peters’ family to determine if this was true. Since the DNA results did not match between descendants of both families, we determined that these two families are NOT connected. 


Next Steps: 

Our next steps would be:

  1. Continue looking for the Martha Peters family represented on the 1880 Oregon Co., MO census report. She was in the right place at the right time and is a good candidate to be the mother of my ancestor, William Edward Peters.
  2. Continue combing through the DNA results of me, my mother, and grandfather to look for matches with Peters in their list of surnames and connections to Oregon or Howell Counties, MO.
  3. Investigate the X matches of my grandfather, since William Peters has been identified as one of his X ancestors.
  4. Apply this exercise to other "suspicious" ancestors: It would be nice to find some descendants of Wilzey King’s family to test and compare to as well. This is another line that I am hypothetically connecting my line to, but am not quite sure of.

Sources:

[1] 1880 U. S. Federal Census, Oregon Co., MO, p. 330C, Martha Peters; digital image, Ancestry.com, (http://www.ancestry.com : accessed 14 June 2011); citing NARA Microfilm publication T9, Roll 707, FHL Film 1254707. Martha Peters, 35 yo (b. abt 1845), TN; William Peters, 7 yo (b. abt 1873), MO, both parents born in TN; James W Peters, 3 yo (b. abt 1877), MO, both parents born in TN; Ella Peters, 2 yo (b. abt 1878), MO, both parents born in TN.
[2] Sources providing information that William’s parents were born in IL, MO, or TN include the 1900-1930 census reports for William Peters and the death certificate of William Peters, 1948.
[3] 1880 U. S. Federal Census, Sebastian Co., AR, p.669C; Samuel Martin; digital image, Ancestry.com, (http://www.ancestry.com : accessed 14 June 2011); citing NARA Microfilm publication T9, Roll 57, FHL Film 1254057.  
[4] With FamilyTreeDNA, the threshold for consideration to be a match is a minimum of 7cM of shared DNA and 700 SNPs.