Michael Cooley's Genetic Genealogy Blog
21 December 2018

The Pettit / Mellowes Y-DNA Profile

The eleven issues of the single-surname journal, The Pettit Correspondent (TPC), were published from April 1988 to Summer 1991. Although my personal genealogical objective was to learn more about the early Pettits of Morgan County, Ohio, I made a sincere attempt to be inclusive by publishing whatever Pettit material came my way. As it turned out, that material far exceeded my capacity to publish it.

Many U.S. Pettit researchers have a vested interest in drawing their lineage back to Thomas Pettit (1609-) and his wife Christian Mellowes of Long Island. One of TPC's purposes was to declare, Not so fast. A large number of Pettits, having various spellings, immigrated to the British Colonies, Canada, and the United States, any number of whom can potentially figure into contemporary lineages. For example, Charles Pettit immigrated to South Carolina at age 30 in 1825, and a John Pettit arrived in Virginia in 1703.1 Although speculation can be helpful to research, jumping to conclusions often hinders it and can frequently result in the pursuit of false leads, sometimes for decades, sometimes for generations, which is what happened with my brand of Cooley (a favorite topic of mine).

Every historian and genealogist knows that huge volumes of records have been lost and will never be recovered. What remains will never lead to a complete picture of Thomas and Christian's descendants. However, DNA, especially the Y chromosome, can fill in some of the empty space and transform conjecture into degrees of probability. False leads, such as the bogus Dutch Cooley myth, can be handily discarded and ignored forever.

At least two members of the Pettit DNA Project at FTDNA, with which I have no connection, claim descent from Thomas and Christian. We wonder at how accurately the lineages are drawn but it's encouraging to see matching results. Among the matches is a tester who claims descent from Henry Pettit (1584-1614) of Essex, England, a man often described (but not proven) as Thomas's father. And there are several other matches to this Y-DNA signature, many of whom have spent decades searching for the elusive Pettit/Mellowes tie-in. Among them is a descendant of Joshua Pettit (1734-1786) of Essex County, New Jersey and South Carolina, as well as a descendant of Elias Pettet (-c1789) of Westmoreland County, Pennsylvania, the progenitor of my own Morgan County, Ohio Pettets. Although both of these lineages are well-established, there has never been evidence to connect them to one another. DNA has done that. For the purpose of this study, it is satisfactory, at least for the time being, to refer to this group in its entirety as the Pettit/Mellowes group. But once all is said and done, we could end up with a different designation.

Short Tandem Repeats (STRs)

The first lesson, always, is that the Y chromosome carries the male sex gene; it comes to the egg via the sperm. Because only a man can have a Y chromosome, it doesn't undergo the recombination found with chromosomes 1-22, the autosomes that many of us have had tested at 23andme, Ancestry.com, or through FTDNA's Family Finder product. That means the Y is a virtual clone: it passes from father to son, generation to generation, over tens of thousands of years. This process creates a near-immutable thread that is handed down for millennia. Any alteration to it is due to the occasional and relatively rare mutation.

There are two types of mutations genetic genealogists study — Short Tandem Repeats (STRs) and Single Nucleotide Polymorphisms (SNPs). Short Tandem Repeats are just that, short repeats of genetic letters that sit near one another in a segment of DNA. For example, at a marker called DYS449, I have 34 repeats of TTTC, which makes me an outsider with everyone in my matching group, including my father. Each of them has 33 repeats at that position. But the overwhelming number of matches I have to the group makes it clear that I belong — even to my father, which was a relief.2

DNA surname projects begin with STRs. They're desirable because they typically show a discernible degree of mutation within a relatively few number of generations. These variations can differentiate between lines that emerged during the genealogical timeframe. For example, this Family Tree DNA chart illustrates that two men having the same surname and having up to three differences among 37 STR markers are almost certainly related. Still, it must be remembered that biology is not rocket science. That is, many sciences, Newtonian physics for one, are highly predictive. Biology is much less so. Although we can hit the bulls eye on Mars, we can't predict what evolution has in store for us. Determining biological relationships, which is precisely what genealogy is about, has more to do with trends than with exact quantification. For example, I have a mismatch of six out of 37 markers to a Cooley with whom I'm related — we're descended from sons of John Cooley (c1738-1811). But FTDNA's chart instructs, Not so fast, and suggests that more research is needed. In fact, like many historic records, STR markers cannot provide the whole story. They can, however, point the way, to show a trend.

I'm not an administrator for the Pettit DNA Project and cannot speak for it. But I've worked with several Pettit testers off and on for years, and with Pettit researchers more than three decades. Seven Pettit/Mellowes testers have shared their Y-STR results with me for this study. Below are the first 37 markers, contrasted against my own to illustrate that, even though I'm likely descended from the same Pettit family, I'm not even remotely related patrilineally, meaning through the father-line. (See the discussion on the M173 marker below.)

Patterns can be difficult to determine in STRs, but we happen to know that the last two listed Pettits are uncle and nephew. The value of 40 at the 36th allele, then, makes sense and indicates that the uncle's father likely had that value. But we don't know how far back it goes. The prevalence of 39 at that position is a strong indicator that that's the ancestral, or modal, value, the value that likely belonged to the Pettit/Mellowes Most Recent Common Ancestor (MRCA). Perhaps that was Thomas; perhaps it was someone else.

We don't know how kits 129959 and 493803 are related. The matching 13 at 9th allele could represent a pattern or it could be simple coincidence. Because STR values can go both up and down, it's impossible to make predictions. Separate lineages can have similar values simply because that's the very nature of STRs — fickle, even volatile. Even the single mismatches, 18 and 12 near the middle of the diagram, can be considered only as one-offs. Until matches with known relationships show up, these are mere anomalies. It's only through comparative analysis that meaning can be gleaned.

Single Nucleotide Polymorphisms (SNPs)

I like to compare STRs to shifting sand dunes; you can't count on them staying in the same state for long. The more time that passes, the greater the landscape differs. SNPs, single-letter mutations that occur at specific positions on the DNA, can be compared to mountains. Genetic genealogists are especially on the lookout for the Mount Everests among SNPs — that is, SNPs that have shown, or will likely show, great longevity. SNPs that appear inside STRs or in the centromeres of a chromosome — the area that links the two halves — aren't good candidates. These are regions that regularly change and have comparatively little permanence — the Sahara Dunes of DNA. We like SNPs that reside in stabler regions of the Y chromosome. Here's why.

Imagine that a father hands a string of identical beads to each of his sons. After three to five generations or so, a descendant adds another bead to the string and passes that down. A few generations later, there's an another addition and so on for, say, a thousand years. It's a broad analogy, but that's essentially what happens to the Y. As noted above, it doesn't go through recombination. Except for the occasional mutation, the Y chromosome is passed down intact.

There are more than 57 million such "beads," or bases, on the Y. (The scientific name is nucleotide, each of which consists of one of four possible molecules, abbreviated to A, C, T, and G.) We can roughly judge the age of a mutation — the approximate time the "bead was first strung" — by determining its distribution in the population. The distribution of the mutation U106, also known a S21, is shown in the graphic from Eupedia. YFull.com estimates that it's close to 5,000 years old. The R-U106 Haplogroup Project at FTDNA makes it about 5,100 years old. In other words, about the time the pyramids were being built, a man in Europe was born with a small mutation at position 8928037 on the Y chromosome. The little gamete (sperm cell) that could had a molecule of thymine (T) at that spot instead of the usual cytosine (C). Our Pettit/Mellowes clan has that same mutation.

Now, let me make this clear: There are no medical or physical traits associated with this or any single-base marker. To provide a gene with the sufficient amount of instruction to make its assigned protein (collagen or keratin, for example), thousands, even hundreds of thousands, of bases are required. The smallest gene is said to be 76 bases long and the longest has 2,304,637 nucleotides. The MB gene on chromosome 22, for example, consists of more than 30,000 bases and is responsible for making myoglobin, the protein shown on the left. A single A or G isn't going to do that.3

Of course, a 5,000 year-old marker isn't going to tell us much about our Pettit origins. U106 came to the British Isles via any number of routes. There's been much said, for example, about certain rare markers found in Britain that likely arrived with the Vikings. But we see from the above map that U106 is as much Scandinavian as it is Franco-Germanic. Genetics is just a piece of the evidence presented for evaluation.

Beginnings of the Pettit-Mellowes SNP Tree

It's the more recently-evolved SNP markers that we're interested in. To that end, one of our Pettit clan, kit B388754, has had ten million and more "beads" on his Y chromosome sequenced by means of FTDNA's Big Y test. The results place him as a direct descendant of the 5,000 year-old U106 man — the man who was first born with that mutation.

The first thing to point out is that the public SNP trees at both FTDNA and YFull.com show that U106 is a single-SNP haplogroup. When it was discovered years ago, there were likely dozens of SNPs associated with it. They peeled away over the years as new testers came on the scene and U106 descendants were found to have only some of them. In other words, new subclades were found. Z2265, the next listed on the diagram, is just one of many sibling clades. And interestingly, both the downstream SNPs from U106, Z2265 and BY30097, are also single-SNP groups. These were rare just a few years ago. And the breaking down of these haplogroups will one day happen to FGC3861, the next downstream subclade. Some of those seven SNPs will be found to belong to separate lineages. To try another analogy, genealogists often understand families the same way. Until a researcher knows more, my maternal grandparents' family might be listed as a small single unit with one known child. In time, the family is recognized as having three children and eight grandchildren. Just as that tree has grown, so will the worldwide SNP tree.

And so it goes — SNPs peel away from a group as new testers and additional information comes along. On the right is an example of the results from several tests by another group I work with. A Hackett tester, the first to test, was found to have 18 novel SNPs. Additional testing slowly spread those 18 among four separate groups. YP4491 is my tribe of Cooleys and the others are distantly related Cochrans. Unfortunately, four years later Mr Hackett hasn't had any matching Hacketts. A properly selected tester would tell us exactly which markers his John Hackett was born with in 1746. In the same way, one day we'll determine the specific genetic profile for Thomas (or ?), as well as for some of the separate lineages from him. The new markers will serve as baseline from which we can approximate relatedness.

It boils down to this: the more Y-SNPs any two men share, the closer they're related. Buried in my Y chromosome, for example, is the SNP M173, the most recent marker I have in common with the Pettit/Mellowes group — and it's about 25,000 years old! In other words, our species was still deep in the Ice Age when a man was born with this mutation. (Position 12914512 morphed from A to C.) One of his sons ancestored the Cooleys and the other became the long dead founder of a lineage that resulted in our Pettits. So much for cousinship. We're all related.

For small numbers of terminal SNPs, like the Hackett group, I recommend that potential matching testers have the SNPs individually tested at Yseq.net for $18 each. But for larger blocks like the present one, it's much better to find one or more Big Y testers. FTDNA.com has an excellent sale now until the end of 2018.

A Pettet Aside (note the spelling)

A book for all Pettit/Mellowes descendants would be quite large. Until recently, I never considered I might be among those listed. For decades, my best guess was that my Pettets of Morgan County, Ohio were not of the Pettit Mellowes cohort. But kit #129959 has proven me wrong. Still, after more than a forty year search, I've neither been able to place my Joseph Pettet (c1816-c1888) directly with any family, nor have I found a suitable tester. To make it even more challenging, there's some thought that Joseph was illegitimate. Autosomes, however, are coming to the rescue.

Autosomal DNA are those parental chromosomes 1 through 22 that recombine with one another to make the unique gametes that become each of us. For example, about half of your #1 chromosome comes from your dad and the other half from you mother. (This, again, is what sets the Y apart from the others: 100% percent comes from the father and women lack it altogether.) I was astounded when I found that my dad has matching segments on chromosomes 4 and 12 to a double Pettet descendant via Jonathan Pettet.

Nancy Pettet, Elijah's daughter, is known to have had an illegitimate daughter, Martha Pettet, in 1818. (Martha married Samuel Pettit — note the spelling.) Nancy married her cousin Plummer Pettet in 1825 and had three sons, including Jonathan. After Plummer's death, she married her widower brother-in-law, Hugh Riley, in 1834. Evidently, Nancy was someone who revered family and wished to keep it together. To make matters more interesting, Hugh Riley, with his wife Nancy Pettet Pettet Riley and her son Jonathan Pettet, relocated to Putnam County, Missouri where my Joseph Pettet resided and raised his family.

Thomas's own nine children have been slowly revealed over the last several decades. There's still one vacancy for a previously unknown male, a man who died by 1888. My Joseph Pettet is the only eligible male not already placed in the family and fits to a T. One day we'll find the two or three unique SNPs that Thomas Pettet of Ohio was born with. If Joseph was Thomas's son, one of his descendants will have those markers — and a certain number of our newly-found 16 SNPs. If not, then Joseph was almost certainly Nancy's son through one of her innovative means of motherhood. Bogey and Ingrid might always have Paris, but by one means or another, I'll always have the Pettet autosomes.


That's a lot to chew on. In some contexts sixteen of anything is an embarrassment of riches. Not so with a 16-SNP haplogroup like ours. In point of fact, it's only the first step toward determining how the Pettit/Mellowes clan hangs on the global SNP tree. I suppose the holiday season is an appropriate time to have found some new ornaments to hang. But a pared down DNA print will be an important piece of evidence toward placing our guy into the right family, the correct birth location, and the appropriate time.

I've written a number of case studies within this blog. If you're confused by this, it might help to read two or three more articles. No matter the surname, this process always works in the same way. And I'm always available for questions. After all, the solution to this puzzle will help me solve my own longstanding Pettet conundrum.

1 Michael Cooley, "Some Pettit Immigrants," The Pettit Correspondent April 1989, 55. A text version of TPC can found at The Pettit Page, http://ancestraldata.com/PETTIT.

2 https://yhrd.org/tools/marker/DYS449.

3 A single SNP mutation can, however, have devastating consequences if it falls in the middle of a protein-coding gene, such as BRCA1 or BRCA2, producers of proteins that help repair damaged DNA.