5/5 ordered
5/7 shipped/
5/10 received and activated and shipped back,
5/14 Ancestry started processing
5/20 to my amazement my results posted.
Not sure one could expect better service then that. The results format is different from the way 23andme and Family Tree DNA presents them  but the way it integrates with  ancestry.com  trees has incredible promise. I had two 4th cousins and the rest more distant. With the low price,fast turnaround and millions of customers I expect the AncestryDNA database to grow rapidly. Lots of things I like include click-able surname lists. I already see that Taylor is a common surname among my matches. Cant wait to see such enhancements as name search among matched trees.

The vast majority of my matches have linked trees , of those that don’t have a linked tree  many allow me to select which of their trees I want to compare. I actually like the ability to select my matches  other public  trees and will write Ancestry to include that option even if a match has linked a tree. The thing I like the most is the openness of the match information. The vast majority of my matches have links to Ancestry.com public Trees this makes it much easier to find matches in common.  23andme by contrast keeps match information ridiculously secret.  While we cant download our results at ancestry.com their fast customer service and openness of results  linked to Family Trees makes be believe their new service will be a  success.

Often times in genealogy research we are faced with the challenge of connecting two people named in two separate  events to the same person. This was true in the case of Philip J. Taylor the sailor who was on the USS Shenandoah that visited Boston in 1953 and the Philip J.  Taylor who was the son of Adlai Taylor who lived in Tell City Indiana.  Family members may know Philip Taylor served in the Navy but I have discovered that distant cousins have no  idea what ship he was on. In addition to my own research needs it became clear to me that if I expect Taylor cousins to help me   I need to convince them that the two Philip Taylors  are the same person so it became essential to my research to verify they were one and the same person. This was the week that I finally was able to connect the dots and here is how I did it.
First we know from my birth mother that Philip J. Taylor was in Boston the Summer of 1953 because of her verbal statements to me and the  piece of note paper with his name and his  ships address showing he was  on the USS Shenandoah. We also have the Navy ships logs showing Philip J Taylor was assigned to the Shenandoah on 11/18/1952 and we have logs and historical documentation that the Shenandoah  was indeed in Boston late July early August 1952.    I have an abstract of an official public  Veterans  record that shows the Philip J Taylor of Tell City  enlisted for his first tour in the Navy November 17,1948 and was released November 17,1952  He reenlisted November 18,1952 and continued to serve until November 15,1958. The fact that the Philip Taylor in Tell City  who  reenlisted November 18,1952 and the Philip J  Taylor who was based on the Shenandoah was assigned to that ship on November 18,1952 appears to be exactly the kind of proof that was needed to show they are the same people.

This is a perfect example of the baby steps and mini successes that make up this journey of the search for my roots, it is simply another piece of evidence to help me connect the dots.

Virtually all genealogy software programs and services such as Ancestry.com allow you to export your genealogy information as a GEDCOM.  GEDCOMS  are universal files that allow easy exchange between software packages and utilities. With a GEDCOM it becomes easy to compare family trees if you have the right tools.  Gedmatch.com allows to to upload your gedcom and compare it with other publicly uploaded GEDCOMS.  GEDMatch.com has even recently added the ability to compare GEDCOMS of your DNA matches.  One disadvantage to GEDMatch.com is not everyone is comfortable to upload there genealogy to a public database.  There are ways to purge all personal data from the files before you upload them , even so many are still hesitant.

The good news is there are also utilities that one can run on their own computer to compare two GEDComs  for common individuals.  Running these utilities on my own server eliminates many privacy concerns  so I hope  my DNA matches will send me their GEDCom file  and I will be happy to see if  I can find our common ancestor.

Even though I am adopted parts of  my family tree have been confirmed by traditional paper trails and DNA matching so I encourage my matches to participate in our mutually beneficial research.  If you need help sending the GEDCOM or  if you have further questions or concerns please do not hesitate to contact me at mark@capeflier.com

Some matches  were willing to confirm  without a known paper trail but others  didnt really understand the benefit and were not willing to confirm without that paper trail.  On  April 3 Family Tree DNA changed this limitation and  you can assign “distant cousin” and even though its only pending you can then use the “In common with” filter.  So If I match Bill Bop and my match also matches Bill Bop he  will show up as a match in common.   This will be a huge help in triangulating results. Lets say we find a paper trail ancestor in common in all three of our family trees, that helps confirm we have the correct common ancestor.

I am not sure how my matches will feel about this but don’t really see the harm in assigning “distant cousin” after all they most likely are in fact a distant cousin we just dont know how yet.  If we do establish a specific known relationship FTDNA also now allows us to change the relationship from distant cousin  to for example 3rd cousin.

For now I am getting the list of  matches in common pasting into excel and then using the assign button to change known distant cousin  back to blank.  Time will tell if my matches  get notified of that brief period of known cousin-ship, if they do Im sure Ill hear about it and will post an update here. If a match wants to keep the known designation I am happy to confirm their request so I don’t see how we can lose.  I do wish we could see matches in common without having to use the assign but this is a great new workaround.

This is an example of a Matrix of Matches in common created with Excel 2010

I am so impressed with GEDMatch.com and it’s sister site ancestor-projects.com that I make regular donations in the hopes that it will continue to be offered for free. I hope others that find the site useful will do the same.

Condensed list of the Ancestors of Owen Kent Cottrell  of Cement City, Michigan

Generation 1
Owen Kent COTTRELL, MI.
Generation 2
William Claude COTTRELL and  Lucy Frances TOMAN,  OH
Generation 3
William COTTRELL,  Eva Myrtle GIBSON, OH
George Leon TOMAN, Mary Florence DICKEY OH
Generation 4
John B M COTTRELL,  Sarah Catherine SPILLERS, Darke County OH
Isaac Newton GIBSON,  Sophia SCHELL, OH
Philip Sylvester TOMAN,  Susan Mary NEWBAUER.  OH
Peter DICKEY, Rosanna BENDER, 28 Jun 1895 in Rose Hill,Darke,Ohio,USA.
Generation 5
Asher R COTTRELL, Sarah Ann SHANER. OH
William SPILLERS, Catherine KINGER.  OH
John GIBSON, Margaret ROSE, OHUSA.
George SCHELL, Margaret Rebecca BOOCHER,  Ohio,
Joseph TOMAN , Sophia GRUBER.
Jacob NIEBUHR (Germany) , Mary Elizabeth LORENZ. England
Jacob DICKEY, Holmes, OH, USA. ,Caterine MARKLEY,OH
John Frederick BENDER , Wurttenburg, Germany.  Christina Magdalena MILLER, OH
Generation 6
Hiram COTTRELL, VA,OH ,Elizabeth REAVES. VA,OH
Adam SHANER,  Elizabeth FIELDS,  Virginia,
Peter Julius SPILLERS,SC,IA ,Susanna COATS, OH
Isaac GIBSON. OH , Married unknown possibly TREGO of Honeybrook PA
Hopkins ROSE,PA, Susana CONNER , PA OH
Jacob SCHELL, Lancaster, Pennsylvania,  Elizabeth KALER, (KOHLER) PA
John BOOCHER PA,  Elizabeth MILLER, PA,OH Samuel Henry GRUBER, PA  Mary Ann SHULTON PA
John DICKEY, PA,OH  Ann Maria DIETZ.
Peter MARKLEY  Mary MILLER. OH
John Jakob (Mueller) Baden-Wuerttemberg,Germany.Anna Marie BUSH  Germany.,OH
Generation 7
John COTTRELL VA , TN Sarah BIRDWELL VA,TN
Asher REAVES, Diana MILLER. OH
George SHANER PA , OH  possibly married Shin BONE Native American   OH
John FIELDS.  VA (Now WV). ,JANE Unknown
William SPILLERS SC/OH, Sarah REAGON. SC/OH
John COATE, SC  Susan BROCK. VA OH
Charles ROSE, RI/WV  Rebecca MULLINER NJ
John CONNER PA ,Sarah TRADER.PA
Andreas Andrew SCHELL , Sophia BECHT.
William KOHLER PA ,Anna ?
Daniel MILLER,  MD,OH , Elizabeth ULRICH PA,MD
Henry GRUBER, PA Anna Margaret KLOPP,PA
George DICKEY, PA Margaret ARNOLD, PA
Jacob DIETZ, PA  Susannah BAKER,PA
Johann George MULLER Germany  Anna Marie GACK.
Generation 8
George BIRDWELL, VA
George REAVES  Wales.,VA  Anna WEBSTER.  Wales,VA
John COATE, NC,  Meary MCCUISTON. PA TN
Elias BROCK,  Ann HOLLINGSWORTH.
Ezekiel ROSE, RI Mary FIDLER.  RI,NJ
Arthur TRADER PA Sarah Sabra CONNER  PA,VA.
Philip Jacob MILLER, PA  Magdalena ROCHETTE MD Kentucky
Henrich Henry Hans Heinrich GRUBER, Germany , Maria Eva EUPHROSINA ROSINA  PA
Johannes Peter KLOPP, PA  Maria Werina BECKER,  PA
Moses DICKEY, PA  Agnes WILSON. PA
George ARNOLD, PA OH  Catherine LEATHERMAN.
Johann Jost DIETZ  Germany, Anna Catherine SCHNEIDER Germany .
Henry or Heinrich BAKER ,OH  Margaret FUNK. Ireland OH

Autosomal genetic genealogy is still relatively new and I think there is still a lot to be learned about the meaning of various degrees of matching.   Most of our research has focused on individuals with shared ancestry that goes back to 1783.  It is a known fact that the effects of recombination results in most 5 and 6th cousins not showing up as matches on Family Tree DNA. Once we examine these results using gedmatch.com and ancestor-projects.com we discover fascinating genetic connections that simply can not be ignored.   We may now have a match that will help us connect all 5 cousins.    There are still some mysteries but we are getting closer to solving them.  I am very grateful for everyone’s help and I am hoping 2012 is the year we get the answers we have been looking for.

To those who are interested here is the process using Windows and Excel 2010:

• In your ancestor-projects project display the matrix showing matches with each other
• On the browser select all
• Copy this data to the clipboard
• Open a new blank sheet in  Excel use the paste special command to paste as unicode text
• delete the first few rows so the first row is the kit number labels
• Make note of the last row with data ie 391
• Move the cursor to the second row  one column further then the last column  to the right with data in the second row (the first row should be the matrix row of kit numbers)
• So in my example column  oc row 2
• Enter the Count  function using my example , make changes to make it work for you   =COUNT(C2:OB2)
• Then copy this function to all the cells below in that same column
• You now have a column with the count of the number of matches for each kit
• Sort this column from largest to smallest  being sure to expand the selection
• There you have it your results sorted with the kits with the most matches on top.

Want to have even more fun
To detect matches in common with another kit

• Copy just tester 1 and tester 2 to row 1 and 2 on their own sheet
• in row 3 use =COUNT(C1:C2)
• Copy that to all the other cells in row 3
• This will populate each  cell in row three with either 0,1,2 depending upon if you match 0,1 or both.
• To determine the number that match both use the COUNTIF function in the last cell
• COUNTIF(C3:OB3,2)

This basic idea works with 3, 4 however many kits you want just change the 2 in the COUNTIF function to the number of kits you are comparing.

The purpose of this  post is to discuss a unique way of looking at the data to see if we can draw any conclusions.

So here is what I did, using the grid in ancestor-projects “Comparison matrix showing relationship between all project participants”  in Lost Colony/Melungeon Lore Project I counted the number of matches with other participants. In other words how many other kits predicted a match within 8.1 generations . Using  my kit as an example I matched 262 of the other kits out of a total of 391. I then sorted the list from most number of matches 281 to least 1.
It is important to note that the kit with the most matches or number 1 on the list had 281 matches. The average predicted genrations to my match was 7.13 That happens to be exactly when  my Cottrell ancestors lived in the Kingsport Tennesseee area in what is now Warriors Path State Park.  Interestingly I have two cousins  with  kits in the project with a  most recent common ancestor between 5 and 6 generations ago. One is number 4 with 270 matches and the other is number 33 with 250 matches.

So my cousins and I are all in the top 10%, I am concluding that the odds of us being  clustered in the top 10% if we were not related is significant. In addition all three of us share 199 of the same kits in common. There are some other really interesting aspects of this data. The average number of matches is 82 the median was 73 . Three quarters of the results were less then 91. The number of matches drops below 200 at number 46 and below 100 at 55 .  The clustering of virtually all of the results in the 0-100 and 200-300  is indicative of something that differentiates the two groups but I have no idea what. I want to make it very clear I am not suggesting that those with fewer matches do not have Melungeon Ancestry just that there appears to be a significant core cluster that have ties to a huge number of  fellow project members.

In conclusion I believe that those of us clustered  in the top grouping must have some sort of significant relationship to each other.     I would welcome thoughts about this.

For those who are interested here is the procedure I used

http://capeflier.com/wp/2011/12/16/sorting-number-of-matrix-matchs-in-ancestor-projects-com-projects/

=====

There are currently 351 DNA kits that are participating in this project.   One aspect of these projects looks at how long a strand of un-recombined shared DNA is and then predicts the relationship between the two individuals. This is given as an estimated number of generations to common ancestor. One cousin was the descendant of George Cottrell a son of Hiram Cottrell   and another cousin we believe was   either Hiram Cottrells brother or another son. I believe I  descend from another of Hiram Cottrells sons Asher Cottrell.

This places our common ancestor in the 1780-1820 range or between 6 and 7 generations.  In doing genetic analysis of a common ancestor between my two cousins and myself we would be looking for estimates between 6 and 8 generations. Some variation might be expected but anyone with a match in the 6-8 generations on its face passes  a reality check.    So for our purposes a match is someone with predicted shared ancestry of up to 8 generations.

One of the great tools available in these projects is a matrix that shows who matches who.  So for example out of 351 kits my kit matches 238 of the other kits within 8.1 generations.  This ranks me in the top 10 in terms of number of matches.  The cousin that descends from Hirams Cottrell’s brother Obedia Cottrell matches 244 other kits also in the top 10  and the cousin who decends from Asher Cottrells brother matches 229 out of 351 testers also in the top 10.    It only gets more interesting from here.  One cousin is number 5 in terms of the most matches, I am number 10 and my other cousin is 16.  If these were juts random unrelated kits they would be in a more normalized distribution. In otherwords they would be randomly distributed throughout the collection of kits.  For those who are into such things the median where half fell below an dhalf were above was only 79, the average was only 87 so 229-244 is substatially outside the curve. Kits above 200 are in a distinct minority with only 45 over 200 and 72 over 100.  I am convinced that the reason these three kits are clustered so close together is that share a common ancestor and the only common ancestor I can see is Hiram Cottrell and his parents. Also interesting is all three of us share 185 matches in common . By way of example if I match kit 101 so do my two cousins and this happens with 185 different kits.   The number of matches places all three of us in the top 10% which defies logic unless one considers were are all related to Hiram Cottrell