By T. J. White (with valuable input from Frank Delano Brewer, with his kind permission) April, 2011

______________________

First, a few definitions are in order, so that those unfamiliar with the subject can have some grasp of what is being discussed. My explanations will occasionally simplify on what is currently known, simply so as not to unnecessarily confuse. Also, please note that I have left out and not defined a number of other genetics-related terms.

MRCA stands for "most recent common ancestor". TMRCA stands for "time to most recent common ancestor".

EPA stands for "earliest proven ancestor" (also "earliest provable ancestor").

DYS stands for "DNA y-Chromosome Sequence".

A 'marker' is a short-hand term referring to an individual STR (short tandem repeat). This is quite a different thing from an SNP (single nucleotide polymorphism).

A DYS number or STR, therefore, is a short segment of the male y-Chromosome which normally in every individual repeats itself a number of times in a stretch of the chromosome. The number of times it repeats in a given individual gives us the DYS (or marker) value. Take, for example, Frank's CDYa==37, versus our modal of 36. In that case, Frank's y-Chromosome repeats that particular segment 37 times, whereas the rest of us "Lanier-Brewers" only have 36 repeats.

Those individuals who have the same number of repeats at the same segment of the y-Chromosome are then shown to be rather closely related. The degree of relatedness is arrived at by statistically extrapolating based on the percentage of identical STR marker "matches," and the extrapolated rates at which these stretches of Chromosome mutate over a known period of time, using a complex system of algorithms. This gives us the probabilities we see that one or more of us share a MRCA within x number of generations.

A Haplotype is the genetic signature or blueprint of an individual, expressed by means of the numerical values obtained when certain STRs are tested, at certain specific sites along the male y-Chromosome. A DYS 'marker' and a DYS 'number' (or value) are two different representations of the same thing: the marker represents the position on the y-Chromosome being tested; while the number or value is-as said above-the number of times the stretch of chromosome repeats at that specific 'marker' location on the chromosome.

A Haplogroup is basically (for purposes of this discussion) all those haplotypes which happen to be closely and provably related, divided into lettered group names for convenience. Examples are Haplogroups R1b1b2 (probably the most common one in Western Europe), I1a, J2b, and our own I1d. This last haplogroup, to which we belong, most probably originated in extreme southern Scandinavia, according to Dr. Ken Nordtvedt, a widely-recognized expert on the "I" haplotypes.

The term 'modal haplotype' means the most common tested genetic values for individual alleles--or genes --within any one particular group of related individuals. It shows what the haplotype of the common male ancestor of the particular group would have looked like--in our case, the "Lanier Brewers".

Nucleotide: This is a sub-unit of DNA or RNA, consisting of a nitrogenous base (made up of the chemicals "adenine," "guanine," "thymine," or "cytosine" in DNA; "adenine," "guanine," "uracil," or "cytosine" in RNA), a phosphate molecule, and a sugar molecule ("deoxyribose" in DNA and "ribose" in RNA). Thousands of such nucleotides are linked together to form a DNA or RNA molecule. These base chemicals named above are abbreviated as "A," "G," "T," and "C" for purposes of discussion or elucidation. A "nucleotide" is also known as either a "base pair," or a single gene or allele in any given chromosome. The position on the chromosome is known as the locus (pl. loci).

An STR, therefore, is a short stretch of several base pairs, in sequence--a stretch which happens to repeat itself in the chromosome any number of times, like a person stuttering. According to Frank Brewer, a sequence such as

AAGCCTA mutating to AGGCCTA is a single letter mutation, i.e., a SNP. TTCGGAT changing to TCCGGAT would be on its copying helix [i.e., the other half of the DNA strand]. Together, the double helix would have one mutated base pair, namely A-T mutating to G-C. As I understand it, there are five possible base pairs: C-G, A-T, G-C, T-A, and A-U. Adenine (A) couples with uracil (U) in receiving RNA. Apparently, RNA rejects thymine (T).

"An SNP," Frank adds, continuing, "in contrast to an STR, is a single mutated base pair. Should the A (vis-a-vis your base pair example of A-T) mutate to G-C then that is a single nucleotide polymorphism."

STR Markers are most often used by us amateur genealogists for determining relatedness; they are a quick, handy, relatively inexpensive way of doing so; the more expensive (and elusive!) SNPs, on the other hand, are usually the only thing looked at by professional geneticists.

This is because SNPs are a more precise way of measuring change, or mutation of base pairs, and therefore of determining who is related to whom, and how long ago their lineages diverged from a MRCA. (Although occasionally some geneticists--even the esteemed Dr. Nordtvedt--will argue for genetic change based only on a significant variance or pattern of STR markers. But this is uncommon.)

Is there any relationship between an "STR" and an "SNP"? Maybe not-at least, not directly. This should be fairly clear from the explanation just given above. (This is the amateur in me speaking.) If, however, on the other hand, we see that a particular haplotype (and by extension, a particular set of unique STR mutations) which is widely shared among a large group of related individuals, seems to exhibit a rather strong stability (i.e. resistance to genetic change) over a long period of time, then we can confidently predict that there is a good likelihood that one or more previously untested, unknown SNPs may well lurk hidden in that haplogroup's genome. This is in fact one of the accepted 'rules' for determining a new subclade of a previously-known haplotype: the new haplotype must possess a unique set of one or more STR mutations, and that haplotype must exhibit a strong stability and resistance to mutation over a long period of time. It is because our "Lanier-Brewer" modal haplotype in fact exhibits both of these criteria, that I am enabled to predict that our haplotype may well constitute a new subclade of I1d, possessing one or more new, as yet unknown SNPs which (if discovered and verified) will verify my theory. This is what I earlier meant when I implied that there might well be an indirect relationship between STRs and SNPs. (Though perhaps the presence of the SNP is in fact what accounts for the STR; who knows? As I say, I'm not an expert on this subject.)

Oh, and one last thing should probably be to point out that although I myself have the surname "White", my DNA is in fact practically identical to that of all you other "Lanier-Brewers"; therefore (as I have explained in a separate paper), one of my male ancestors was in fact fathered by a Brewer, of the same line as all of you. I am your genetic cousin, my present different surname notwithstanding.

_________________________

Here follows most of the discussion between Frank Brewer and myself. I had sent him an advertisement from FTDNA (Family Tree DNA) regarding a new DNA test of theirs, asking whether or not he thought the test would be helpful for our research:

Frank: Terry, thanks for the info.

Out of 67 markers tested, you and and I match 65. If we each test and match an additional 44 markers, we would still have two mismatches. I'm satisfied that George Brewer and Sarah Lanier are our ancestors.

Paris Brewer - my distant cousin - and I match 67 for 67. I'm highly confident that our MRCA is Willis Brewer (1799-1848). It may be helpful for new distant cousins (yet unknown) to test the same 67 markers. I see no personal advantage for cousins - who have already tested their 67 markers - to test additional markers; however, it would be advantageous to Family Tree.

_________________________

Terry: It would definitely be advantageous for FTDNA (LOL) ... Got a laugh from me there. Yes--good as always to seek your input. So I think that's how we'll go. Just need to find new suspected cousins.

_________________________

Frank: Yes, additional cousins testing Y-DNA would help, especially if testing 67 markers. Meanwhile, we could mine nuggets from existing records of cousins who have already tested 67 markers.

To date, I have 14 matches at various genetic distances. You may have most of those matches. One name you might not have is "Henry Frank Brewer" (for me a genetic distance of 7). Who is listed under your genetic distance of 7? Do you have "Henry Frank Brewer" listed anywhere?

The FTDNA Project Page for George Brewer/Sarah Lanier Descendants ... .

http://www.brewer-family.org/genealogy/brewer/ftdnapedigrees.php

... lists groups. Although I see names of Willis Brewer descendants, I've no confidence in their alleged connection to George & Sarah. Rather than speculate, one chart listing all names and 67 markers (across the top) may be revealing. Example: two cousins may each have another cousin listed at a genetic distance of 5 but not the same 5. A chart could reveal - at a glance - groups who share identical markers. Such groupings would be scientific rather than speculative.

____________________________

Terry: Yes, one single chart listing all 67 markers is indeed advantageous.

As you may recall, I have already made just such a chart (see attachment), and indeed, Henry Frank Brewer is already listed there. I show him as a genetic distance of 3 to myself (64/67), but he is genetic distance 2 (65/67) to my 4th cousin Monty White, who shares one more mutation with him than I do.

Does this chart of mine fit your bill of being a little "more scientific"? It not only compares all 67 markers of everyone I know of who matches us (Brewer surname or not), but also says what I know of each person's EPA (earliest provable ancestors). This enables precisely the sort of comparison you envisage.

I hope you have as much fun with this chart as I have had! Please let me know how it 'hits' you. I think that, with this chart (or another one much like it), we can start to make some real progress.

Next step is to take this chart and correlate it with the exhaustive survey of Brewer relationships shown in Foy Varner's voluminous Brewer Families of Southeast America. I have that book as a Word file, and can send it, should you not have a copy. Varner's look at all the known Brewers is about as thorough, cautious, scientific, and documented as they come. His work will be the basis for anything we can ever think of doing in the future.

_________________________

Frank: Terry, although it doesn't impact your search, you may find this technically interesting. I descend from Willis from his son, Lewis. Paris descends from Willis and his son, Isom/Isham. Paris and I match 67/67, so we would probably match Willis 67/67.

Daniel descends from Willis and his son, Thomas. Daniel differs from Paris and me by one marker; therefore a mutation probably occurred in the Thomas branch of our family tree.

Terry, your chart helped clarify this for me. I'll try emailing the info to Daniel.

_________________________

Terry: Not "probably match[es]," but "provably matches". That is absolute proof of your ancestor Willis Brewer's exact haplotype, just as in the case of myself and my known 4th cousin, once removed (Roger White): I descend from Wm. Wilson White (1800-1895), and Roger descends from Wright White (1807-1893). Since those two men were known brothers to each other, and since Roger and I match 67/67, that proves beyond all scientific doubt the exact haplotype of the common ancestor of myself and Roger, the man who was known as Jacob "Jake" White (whose father we now know to have been a Brewer).

Monty White, who is also a 4th cousin to Roger, and a 4th cousin, once removed to myself, has an STR mutation at DYS 570 (20 in Monty's case, versus the modal 21, which Roger and I both share). This means that that particular mutation occurred in one of Monty's male ancestors, after the generation represented by Jacob "Jake" White. It could have been Jacob's son Samuel Isaiah White (1810-1893) (Monty's ancestor), or any one of his male descendants, all the way down to Monty himself.

The only way to isolate when the mutation occurred is to do more tests from other males from that specific line; in Monty's case, say, for instance, that we test a known first cousin to Monty, and he does not have that mutation; in that case, we can say for a certainty that the mutation had to have occurred in either Monty or in his father (since the common grandfather of Monty and his first cousin would have then been proven to have not had that mutation).

We could then further refine (isolating the exact generation when the mutation occurred), by testing one or more of Monty's brothers. If they all share that mutation, then the mutation occurred in Monty's father, Eli Hugh White (1890-1958). If they do not share that mutation, then it had to have occurred in Monty himself.

This is only a hypothetical example, of course. The same would be the case in Willis Brewer or among his known male descendants. And of course, we should probably be reminded that genetic mutations are a completely normal, commonplace occurrence: they happen to every human (and animal) family over enough time. They are in fact part and parcel of the evolutionary process itself-evolution having been made observable, or visible to the naked eye, by means of the tool of the genetic test, if you will. Instead of looking at major changes in or between species generally (which normally take eons to occur), we are looking at biological/genetic changes in only the briefest of evolutionary moments, and on practically the smallest scale possible. I don't know about you, but I find this to be a fascinating process to watch and be a part of.

I know this will all make perfect sense to you, once it's explained. This is precisely how further 67 marker tests will help us re-write our family tree. In fact, though, we don't necessarily need to test all 67 markers, in these new test candidates; we really only need to test the one panel where the mutation(s) in question are known to reside. That should be much, much cheaper. (That's assuming FTDNA will permit that sort of limited testing; they may, in fact, not permit it.)

The mutation in question in Monty's case (and in your ancestor Willis' case as well) is in Panel 3 (STR markers 25-36). If we could test only that one panel in all new test candidates known to be related to us, then we would save a lot of money, and still have the results we need. This would be a great boon to our efforts to scientifically re-write our family tree.

No, I think you and Paris are indeed descended from George and Sarah--just as my branch probably is. We too share a rather unique mutation among ourselves (DYS 395S1a=14, versus the modal of 15), and this too separates my branch from the "pack". So this is not automatically grounds for presuming that George and Sarah were not our MRCA. A few of the others also have unique mutations, which also separate them as well. That fact alone should not be grounds for undue concern. Any or all of those mutations could well have occurred in the twelve or so generations since George and Sarah.

Some months back, I sent a copy of this very same chart to Dr. Nordtvedt, and got some useful feedback from him. Among the several things he had to say about our "Lanier-Brewer" results, was that there seems to have been a "mass adoption event" in our genetic family in Colonial North Carolina, and also that several (if not most) of the STR mutations which seem to separate and delineate different branches in our "Lanier-Brewer" group were probably only the result of what geneticists call "convergence"--i.e., mere accidental resemblance, and later, independent (though congruent) developments in a more recent time frame. In other words (said Nordtvedt) most of those STR mutations probably do not describe separate lines of descent from a common ancestor, but rather show a genetic tendency to mutate in those directions in this particular family.

Take my own noteworthy mutation at DYS 395S1a=14. There is another Brewer who, interestingly enough, shares that same mutation. He is the same Henry Frank Brewer you have already mentioned. But does the fact that he shares that mutation with my two "White" cousins and me, mean that Henry Frank is a descendant of our common "White" ancestor, Jacob "Jake" White? Hardly! It is merely a random, happenstance occurrence which is predicated upon a genetic tendency to mutate in that direction, at that particular DYS location, in our genetic family. Ditto with most of the rest of the mutations observable in my chart.

But not all such mutations are the result of "convergence" only; some few do indeed (perhaps) serve to define separate lines of descent from George and Sarah (our MRCA). Take, for example, your own DYS CDYa=37, versus the modal 36: that STR mutation would indeed seem to isolate the descendants of Willis. Even here, though, yet another Brewer not descended from Willis also shares that same mutation; he is Richard Rusty Brewer, descended from Julius Brewer, born 1794 in NC. So that particular mutation you and Paris share may not mean much, after all.

The same is true for my own shared mutation of DYS 395S1a=14 versus the modal of 15. It too would seem to define a separate line of descent. (Though here again, reminding you of the fact that yet another Brewer-Henry Frank--also shares that so-called 'defining' mutation!)

Needless to say, it is very, very tricky to say which mutations are "convergence," and which are the result of a shared inheritance from a common ancestor (and Nordtvedt himself repeatedly expresses this caution). Just when we think we've found one, up pops another cousin with the same mutation, calling our theory into serious question. About the only way, even approaching being 'foolproof', to determine which mutations are convergence, and which are shared inheritance, would be to obtain a dramatically large pool of test subjects, for comparison. And we are far from that stage still.

My genetics friend and guru, Carol Vass, who knows of Nordtvedt quite well (by reputation), assures me that I did well to capture his attention with regard to my DYS Chart and our "Lanier-Brewer" haplotype. "You got his attention," said she, telling me that that was a good thing, and hard to accomplish. Nordtvedt told me that he was adding our "Lanier-Brewer" modal haplotype to his database ("quite an accomplishment," said Carol).

I had written to Nordtvedt mainly to voice my strong suspicion that our "Lanier-Brewer" group, because of no fewer than five (5) separate STR mutations, mutations which serve to separate our modal haplotype from its "founder haplotype" (L22-NuN14), probably had one or more unknown and untested SNPs lurking around in our genome. To my eternal satisfaction, Nordtvedt agreed with me, but added, "Who will you find to test them for you?"--meaning, 'how will your group afford to pay to have them isolated and identified by a lab capable of doing so'?

Good question, no? But whoever pursues it will undoubtedly be credited with discovering and identifying one or more unknown SNPs. Worthy goal, I think. And I was apparently the first to suspect they were there.

Note: I am indebted both to my "genetics guru" Carol Vass, and to my esteemed cousin Frank DeLano Brewer, for helping to educate me on the subject of genetics. They both have my deepest gratitude.