How my AncestryDNA Match stands up in Gedmatch

In my last post, I evaluated one of my "shaky leaf" matches whom AncestryDNA predicted would have a shared ancestor with me. We compared our online family trees and determined that we were connected via the Long/Lang line of Jasper County, Georgia. This match was instrumental in both of our research efforts because conflicting documentation regarding the Long/Lang surname had made it difficult to place our ancestors with the correct set of parents. This DNA match allowed us to confidently place our ancestors within the same family of John and Margaret Lang.

I have to admit, by the time I finished writing this last post, I was pretty hooked with using AncestryDNA. But I know when something seems too good to be true, it usually is. I knew there were going to be limitations with the AncestryDNA service going in to it. And I wanted to put this to the test. Unfortunately, my suspicions were confirmed:  My AncestryDNA match and I were not really connected, genetically speaking,  via the Lang/Long line!

My Pedigree:

My descent from John and Margaret Lang is as follows:
Basheba Long & Washington Phelps
Lucinda S Phelps & Burwell Binns
Milton A Binns & Susan Ann Dupree
John Milton Binns & Perthinia "Pert" Brooks
John Brooks Binns & Blanche Kathryn Hill
B Binns and D Smith - my grandparents
T Smith & M Godwin - my parents
Ginger Smith - me

I have tested myself, both of my parents, and 3 of my 4 grandparents to date. If I received any DNA from my Lang / Long ancestors, I would have inherited it from my Father, who inherited it from his mother, my grandmother, B. Binns. However I tested all of us at FamilyTreeDNA and tested only myself at Ancestry.com. In order to compare any DNA results from AncestryDNA to results from FamilyTreeDNA, they must first be uploaded to a 3rd party website called Gedmatch.com. GedMatch allows you to compare autosomal DNA across all 3 testing companies - Ancestry.com, 23AndMe, and FamilyTreeDNA. It's FREE and easy to use!

What I Did:

I convinced my match to upload her AncestryDNA results to Gedmatch.com. Once uploaded, I compared her results to my grandmother, B. Binns, who descends from our Lang / Long ancestors. I expected to see a good match since my grandmother is closer to the shared ancestor than I am. Surprisingly, there was NO MATCH. Not only was there no match, but my match did not share the same segments of DNA with my grandmother that she shared with me.

Here is a list of segments S.A.T. shares with me:

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
3	25052678	30683436	7.2	1487
7	88445250	91973622	3.0	731
7	117280746	123629411	3.5	933
13	66743336	69809017	1.9	739
18	22684849	25803725	3.3	977

Largest segment = 7.2 cM

Total of segments > 1 cM = 18.9 cM

Estimated number of generations to MRCA = 6.8

Here is a list of segments S.A.T. shares with my grandmother:

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
1	95871096	98378320	1.6	970
2	34843826	37164843	2.6	705
3	47546444	53219643	1.0	801
10	73326487	77694862	2.8	759
12	20794787	22833572	3.1	812

Largest segment = 3.1 cM
Total of segments > 1 cM = 11.1 cM
Estimated number of generations to MRCA = 7.2

There are no overlaps between where she matches my grandmother and where she matches me.


Is this For Real?

If you look at how S.A.T. matches to me, you will see that the MOST that we match on is 7.2 cM on ch 3. Our total matching segments are only 18.9 cM. This is a very low resolution match in my opinion. In order for someone to be considered a "match" in FamilyTreeDNA, they have to share at least one segment of DNA that is 7 cM long - this requirement IS met in this example. The 2nd requirement by FamilyTreeDNA is to have a minimum of 20 cM total DNA shared between two people to be considered a match. S.A.T. and I only share a total of 18.9 cM, so this would not fulfill the requirement to be considered a match in the FamilyTreeDNA system. So is this a real match or not?

Not Maternal..., so Paternal?

Although I was discouraged, I did not let this set back keep me from looking at my match more closely. Although she wasn't a match to my grandmother as I expected, I then compared her to the rest of my family members, including my grandfather and my mother and her father. She DID come back as a match to my paternal grandfather.

Here is a list of common chromosome segments between my match and my grandfather:

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
3	25052678	30677144	7.2	1475
5	50991761	54288881	2.9	846
6	151277980	153997386	4.8	924

Largest segment = 7.2 cM
Total of segments > 1 cM = 14.9 cM
Estimated number of generations to MRCA = 7.0

Here is a list of common chromosome segments between my match and me:

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
3	25052678	30683436	7.2	1487
7	88445250	91973622	3.0	731
7	117280746	123629411	3.5	933
13	66743336	69809017	1.9	739
18	22684849	25803725	3.3	977

Largest segment = 7.2 cM

Total of segments > 1 cM = 18.9 cM

Estimated number of generations to MRCA = 6.8

You can see that my match shares the SAME chromosome segment on ch 3 from 25052678 to 30683436 with me that she shares with my grandfather. That segment is 7.2 cM long. This means that our genetic connection came down to me through my grandfather, not my grandmother, and that our common ancestors ACCORDING TO THE DNA, are not the Langs or Longs, but a different set of ancestors shared on my grandfather's side.

I am happy to have connected with my match on the Lang / Long line, and to have extended my tree out further with the addition of her line; However, I can no longer conclude that "DNA proves that my Basheba Long was the daughter of John Lang" ...because I am a DNA match to other descendants of John Lang....

This is where we have to draw the line between DNA testing and genealogy and make distinctions about what each one can do. My mantra has always been, "If you're going to do it, then do it right!" The bottom line is that you can't really use one research method independent of the other. You can't rely solely on the DNA to give you answers and sometimes the paper trail just isn't enough.

Conclusion:

AncestryDNA is a good tool to use to find connections and to expand your paper trail or family tree. However, its usefulness is limited with regards to DNA because the chromosome data is not made available to the testee. In this example, I found a connection with my DNA match via our online trees and genealogical paper trails. However, when I input the raw DNA data into Gedmatch.com, I realized that the DNA connection, represented by the DNA that is shared between us, was not carried down to me from my Lang or Long ancestors like the paper trail suggested. A side by side comparison of the segments my match shared with me to the ones she shared with my grandmother clearly showed that I did not receive the segments that I shared with this match from my grandmother who is a direct descendant of the Langs/Longs. Additionally, when I compared my match to my grandfather, it was clearly apparent that the segments I shared with my match were carried down to me through my grandfather because he, too, shared those same segments with my match.

Now, who's going to break it to my match that we have to go back to the drawing board and start over???

Related Posts: 

My AncestryDNA Test, part 1 - A look at the Closest Matches

My AncestryDNA Test, Part 2, The Shaky Leafs

My AncestryDNA Test, Part 2, The Shaky Leafs

FGS Report of my Lang Ancestors

This is the second post in a series about my AncestryDNA results. In Part 1, I discussed a 3rd cousin match I had found. I found him by looking at my top matches and clicking on one. He was one of two 3rd cousin matches. I clicked on his name, glanced at his tree which contained only 9 people, and immediately identified the connection. Even though I knew right off the bat what our connection was, AncestryDNA did not provide a Shaky Leaf "Hint" because he had not added enough people to his tree (remember I said he had only 9 people in it?). I'm sure that once he starts adding people to his tree, including his 1st, 2nd, and 3rd great-grandparents, whose names I gave to him, a shaky leaf will eventually appear.


The Shaky Leaf and How I found My Next Match

My DNA matches are sorted by relationship by default, with the closest relationships at the top of the list. Although a relationship is predicted, it is not guaranteed that a connection will be found and I will be able to place my match on my family tree. Another way to look at my valuable matches, is to sort by "Hints" represented by, you guessed it, those shaky leafs! In order to see my "Hints" I clicked on the "Hint" filter at the top of the page. This resulted in two pages of matches - about 100 total matches. I found my match to S.A.T. on page 2.


The Relationship Report

When I clicked on S.A.T.'s name, I got a report that told me the following:

• Our predicted relationship was Distant Cousins
• There was a Possible Range of 5th - 8th cousins
• There was a Confidence of Very Low

This looked pretty dreary at first. But then I scrolled down. This is where the Ancestor "Hint" came in. I was presented with a Relationship Report showing me how I was related to my match and who our common ancestor was based on that relationship. According to this relationship report, our common ancestor was John Lang, my 6th Great-Grandfather.

This comparison was run by the Ancestry.com servers (in the background - Ancestry.com people say it's an algorithm). For every match that has a tree, the servers run a side by side comparison between their tree and mine and when they find someone in both trees that looks like the same person and it is a direct ancestor of both people, they determine it to be a "shared ancestor."

Relationship Report with Ginger Smith on the left and my Match S.A.T. on the right

Finding DNA evidence linking my 5th great-grandmother to John Lang has been pretty instrumental to my research. And evidently, it was instrumental to my match's research as well, so she was pretty excited to learn of the match. We both had conflicting evidence about our Lang/Long ancestors that had yet to be resolved.

My Research Problem - the Lang/Long Conundrum

My 4th great-grandmother, Lucinda Phelps, and her husband, Burwell Binns, moved with their families from Talbot County, Georgia to Drew County, Arkansas about 1856. Lucinda's mother, Basheba Lang/Long and step-father, Green Berry Jackson also moved with them. Burwell died there in 1866 and his wife Lucinda died shortly after in 1870. Both Burwell and Lucinda were buried in Beulah Cemetery in Veasey, Drew County, Arkansas on land they donated to the Beulah Baptist Church.

Basheba and Green Berry died after Lucinda and Burwell, however, their final burial place is unknown at this time, but presumed to also be in the Beulah Church Cemetery. There is a piece of paper that someone tacked up behind the glass at the Beulah Church Cemetery entrance that talks about Lucinda's parents, including her mother, Basheba LONG:

Here is a transcript of this informative paper:

"Basheba LONG married Washington Phelps 21 April 1816 at Randolph GA. We have knowledge of one daughter, Lucinda Phelps, being born to them. 

I. Lucinda Phelps (born 22 Oct 1818 in Jasper Co GA) Lucinda married Burwell Binns II in GA on 29 Oct 1835. The had ten children and this is the Binns family who settled in the Beulah Community of Drew County. The children were: Milton, Sarah, Elizabeth, Lucinda, Matilda, Edna, Mary Jane, Christopher, and Burwell III.

Basheba Phelps married Green Berry (sometimes called Greenberry) Jackson 21 March 1823 in Jasper Co GA. Court records show Green Berry Jackson as being appointed guardian of Lucinda Phelps (dau of Washington Phelps) sometime after 1823. We have knowledge of 4 children belonging to Green Berry and Basheba Jackson and the names of 3. By the time Green Berry and Basheba Jackson got to Arkansas, Basheba was spelling her name Barshaby. We don't know when this family arrived in Drew County; but they were here by 1857. Children of Green Berry and Barshaby Jackson are: ...."

In this write up, Basheba's last name is "LONG." However, on Basheba and Washington's marriage record recorded in Jasper County, Georgia, she is listed as Basheba LANG.

Washington Phelps & Basheba LANG Marriage Record Jasper Co GA, 1816

Here is the transcript of the marriage record filed in Jasper Co., Georgia:

Georgia       ))

Jasper County ))

     I certify the within named persons Washington Phelps and Bersheba Lang were duly Solemnized in the holy bond of wedlock by me this 21st day of April 1816. 

                                                              Luke Williams, J. P.

Robert Robey, C.C.O

So far, we now have two conflicting pieces of documentation for Basheba's surname. The paper posted at the cemetery said her last name was LONG and the marriage record said her last name was LANG.

But I wasn't the only one with conflicting documentation. My match had run into similar issues with her family.

My Match Has the Same Lang/Long Conundrum

According to my match, her 2nd great-grandmother was Temperance Lang who was born 1850 in Stone Mountain, DeKalb Co., GA and died 1928 in DeKalb Co., GA. On her death certificate, the informant listed her father's name as John C. LONG, however, on every other record she encountered, including the death certificates of Temperance's siblings, her father's name was listed as John C. LANG.

With DNA testing, we were able to confirm that our two lines were related and that her John C. LANG was the sibling of my Basheba LONG. I had previously connected to another sibling, Lydia Lang via a DNA match as well, so I knew we were on the right path.

In this case, DNA helped to solve two mysteries and hopefully put the Lang/Long conundrum to rest.

As you can guess, I will be looking for additional documentation on my Basheba Long/Lang in order to determine if she changed her name, went by both names, or just didn't bother correcting the county clerks who wrote up various documents for her.

I would love to hear if DNA has helped to resolve any of your name conflicts. Please leave a comment below or email me directly.

Genealogy Goals for January 2011

So I didn't get as much done over the Christmas break as I had hoped. But I did start a couple of things that I hope to finish in January. I thought I would share them here:

1) To scan all of the genealogy documents that my cousin Richard sent me on the Binns and Brooks sides of my family and enter the information into my genealogy database

2) To enter all of the Binns, Anthony, and Pye names references from the Wilkes County Georgia books I copied from the Library into my excel spreadsheet. Determine the major Binns family lines in Wilkes Co. GA

3) To read Christine Rose's Courthouse Research for Family Historians: Your Guide to Genealogical Treasures book

4) Get back to working with some cousins on James Godwin Sr and Jr (not necessarily father and son) of North Carolina, including putting all relevant documents online on the Wiki site

5) Continue evaluating some Family Finder DNA test results of some Godwin relatives

Here's a toast to typing up loose ends!

Photo courtesy of D Sharon Pruitt