Michael Cooley's Genetic Genealogy Blog GEN • GEN
21 June 2017

Cooley CF01 DNA Updates

New DNA test results have been posted for the CF01 Cooleys and Whitfields. No single piece of data is dramatic but, taken together, they're interesting.

This simple diagram from article 36 illustrates the relative cousinship between the three CF01 groups: the Cooleys having an unknown MRCA (Most Recent Common Ancestor) in Pennsylvania, the descendants of John Cooley (c1738-1811) of Stokes County, North Carolina, and the descendants of William Whitfield (1751-1835).

I reported last time that three descendants of Henry Edwin Whitfield (1819-1884), William Whitfield's grandson, have matching STRs to the Cooleys of Stokes County. One of those three, FTDNA kit #520597, went whole-hog and tested ten million bases on the Y chromosome, a test known as the Big Y. We found that he has the four YP4491 terminal SNPs — previously known only to belong to the Cooleys — and two new SNPs, A14495 and A14496, that had come down to him somewhere in his lineage.

This is good but incomplete information in that it provides the genetic Y-print (which is very Cooley-like) for Henry Whitfield but not necessarily for his grandfather, William. A tester from a collateral line was needed before a determination could confidently be made. Sure enough, another match, a descendant of John Toler Whitefield, one of William's sons, soon came along. This well-known genealogical relationship has now been confirmed through genetics:

Just as interestingly, John Whitefield's descendant is negative for the two additional SNPs discovered in the Whitfield Big Y. This means that William Whitfield had no terminal SNPs beyond the Cooley YP4491 block of four. His SNP composition is identical to John Cooley's.

The last bits of data from the Whitfield Big Y test, 500 STRs, have now been processed. Whereas SNPs (Single Nucleotide Polymorphisms) are rock solid over hundreds, even thousands of generations, Short Tandem Repeats (STRs) — short strings of genetic letters repeated X number of times — tend to be rather fickle, which is precisely why they're used for FTDNA's surname projects: differences can show up after only a few generations. It's these differences, the genetic distance (GD), that leads us to believe that the Pennsylvania group of CF01 Cooleys is somewhat more distantly related from the Stokes County Cooleys than are the Whitfields. The last half of article 22 is devoted to describing those STR differences. But suffice it to say that significant and consistent differences are present at markers DYS464b and DYS376.1 The two groups of Cooleys share a common ancestor but have verifiably different Y-prints. However, at 37 markers, Whitfield is a GD of 0 from the North Carolina Cooley STR modal. What about at 500 STRs?

Three CF01 members, including myself, have done the Big Y. In this table, A and B are Cooleys, C is Whitfield. The numbers represent the GD between one another:

This is virtually a draw. No matter how paired, the GD is about the same. We can find no mutation difference, SNPs or STRs, between the Cooleys and Whitfields. It's almost as though John Cooley and William Whitfield were the same man. Of course, they're genealogically very distinct. They may have had much the same origins, but their life's paths were separate.

Family Finder clues

I write little about autosomes, those pairs of chromosomes (1-22) we inherit equally from both parents. Unlike the Y chromosome, they don't clone, they recombine. We get half from each parent, a quarter from each grandparent, about an eighth from each great-grandparent, and etc. Whereas paternal fifth cousins will have virtually the same Y chromosome, autosomal fifth cousins will share only a small number of genetic bits — and number is the operative word.

Autosome matches are measured in centiMorgans. A parent and child will share about 3500-3700 cM; first cousins share between about 550 to 1150 cM; and fourth cousins zero to 54 cM. A high count provides a large degree of certainly about the relationship. But so much DNA is "shed" over the lineage from any one distant ancestor, that it's difficult to make a solid prediction.

Personally, I generally don't explore predictions much greater than third cousins. But even if one wishes to take a fifth-cousin match seriously, how is it determined which of the thirty-two pairs of 4th great-grandparents is the match? Most of us don't even have a complete seven-generation genealogy. If such a DNA match is legitimate, which of the blank spaces on our charts does it represent? And distant cousins are often multiply related. Even if they're known to the record, which set of MRCAs is the matching couple — or do we have DNA parts from both? All too often we're at a loss to explain small matches, and if we can't explain them, we can only be consider them footnotes.

Still, two people could share any of tens of millions of potential ancestral candidates. What are the chances that a match exists and you recognize one of the ancestors of a possible sixth cousin? Consider the following:

This is only a cursory look at a few members of CF01. Taken individually, I'm not impressed, but collectively, they seem to dovetail with the inevitable conclusion: John Cooley and William Whitfield were closely related. They were so close, in fact, that there are no distinguishing Y markers among their descendants. We can trace both men to the same region in Virginia, and they are both believed to have been born in England. But what are the chances of finding the genealogical record that explains the connection? Very remote, I think. Our best bet, I believe, is to find YP4491 SNP matches in England. I'm open to suggestions on how best to achieve that.

1 The Pennsylvania group's values for DYS464b match with families of Hacketts, Cochrans, and Staries that split off the SNP tree hundreds of years earlier. (See article 10.) For this reason, I'm tempted to refer to the PA group as "Old CF01" and the NC Cooleys and Whitfields as "New CF01."