The Basics of Y DNA Testing
We get many things from our fathers. But if you were to construct a list of all the things your father has passed on to you, I am pretty sure STRs and SNPs would not be on your list. Yet these genetic biomarkers have been faithfully passed from father to son over the centuries and they are the essential building blocks of Y DNA testing.
STRs (Short Tandem Repeats) and SNPs (Single Nucleotide Polymorphisms) are two different ways to differentiate between men when comparing their Y dna. Because neither are recombined with the dna of a female through reproduction, both tend to be passed along identically from father to son from generation to generation, and only gradually change, completely independent of one another. As you may know, there are four types of bases found in a DNA molecule: adenine (A), cytosine (C), guanine (G), and thymine (T). Adenine pairs with thymine, and cytosine pairs with guanine.
With STR testing we are looking for recurrent patterns or chains of these bases at specific locations along the Y chromosome (Y-12 locations, Y25 locations, Y-37 locations, Y-67 locations, Y-111 locations, Y-700 locations). For example, "CTACTACTA" would represent an STR value of 3 base pair repeated sequences of CTA at a given location on the Y chromosome. Over time, a descendant of such an individual might pick up another repetition of the sequence "CTACTACTACTA" giving him a value of 4 base pair repeats at that location. Another descendant might lose a pair "CTACTA" giving him 2 repeats. Once these repeat sequences change, they are passed on. Sometimes they revert back, though which makes STR typing an inexact science.
SNP testing is a more exact science as SNPs once changed almost never revert back. For instance, the sequence, GTACAAGA may over time mutate to GTACTAGA as one single nucleotide adenine base (A) mutates to thymine (T). Now all of this man's descendants will show T instead of A at that location. These stable changes can be dated and regionally mapped as additional men show the same mutation change.
At this point you might be wondering how do STR testing and SNP testing compare? If I could liken STR testing to looking at a photograph of a group of men trying to guess who might be closely related from common shared features, then the Y-37, Y-67, and Y-111 tests would compare to seeing the same photograph progressively more closeup so that more shared features would become discernable in the photo, allowing you to group men by shared physical features. With a limited degree of certainty you might have a certain degree of success with this technique especially among siblings and close cousins. Now in point of fact, matching men by STR patterns is way better than sorting them out by shared physical features, especially since STR patterns , unlike shared physical features, do not get influenced by maternal contributions. STR patterns also are more distinguishable and more enduring than physical features. Yet both techniques can be hampered by false positives where shared features may incorrectly suggest common lineage (especially at lower levels of STR testing like Y-37).
Now SNP testing adds learning who definitively shares the same ancestor at various statistically derived points in time. For instance, if all of those men in the photo subsequently did SNP testing and several showed a particular SNP that is known to have developed about 600 years ago (e.g. R-Z16280 is one such SNP), that means those particular men shared the same paternal great grandfather roughly 600 years ago. Very likely you would be surprised to find that at least a few of the men in the photo who appeared to resemble each other didn't share many SNPs. Thus the similarities between their physical features were not genetically related through their father's fathers in recent history. Similarly, you would also likely find that only a few of their STR matches shared their most recent SNPs and not necessarily their very closest STR matches.
One prominent example of what can be learned from SNP testing particular to the McMahon Surname Project can be seen with the SNPs R-A77 and R-Z16280. An unknown male living in Airgíalla (roughly present day County Monaghan and parts of adjoining counties) developed the R-A77 mutation about 1200-1300 AD. One of his offspring subsequently developed an additional SNP mutation in the same region of Ireland about 1400 A.D. Because so many Colla McMahons carry these particular two mutations, we think it very likely that those carrying these SNPs descend directly from the royal McMahon overlords of the time. As Patrick McMahon, our Project Coordinator, has elsewhere written, "There could have been quite a McMahon population explosion (from a single source) over the three Monaghan baronies (Farney, Luchtighe, Dartrey) roughly between the beginning and mid 15th c. Firstly, there was Ardghal (d. 1427) who gave rise to Ruairi Mor (Ur-Ri Oirgialla, d. 1446) and 11 other sons. Then Ruairi in turn had Reamonn (Farney, d.1484), Aodh Rua (Luchtighe, d.1453), Eoghan (Dartry, d. 1467) and 13 other sons, a total of 28 breeding males over a 40-50 year period from one family line over all Monaghan, all exhibiting indistinguishable DNA patterns (then). With so many parallel lines of descent and a lack of documented pedigrees, it is impossible to identify who, if anyone, is of direct descent from any of the named progenitors". Nevertheless, you can now see how SNP testing yields much valuable information beyond what could be learned from STR testing alone.
It has become apparent that some testers feel that testing for low numbers of markers is adequate once a match has been discovered. This can sometimes be a false promise and it is not always easy to explain.
It's generally true that you can 'see' a relationship at 25 markers if you can distinguish it from all the other dross that testing at this level throws up. 25 marker testing tends to produce lots of false positives which can only be resolved with a greater number of tests.The greater the number of tests, the greater the statistical significance. FTDNA considers a genetic distance (GD) of 2 as the significant limit at 25markers, 4 at 37, 7 at 67 and 10 at 111. It then depends on which actual markers are different. Say you have a GD of 5 @ 67 markers which is highly significant; if these differences occurred in the first 37 markers, you probably wouldn't see this relationship at 25 or 37 markers (because of FTDNA cut-off points). Consequently, we encourage new and existing members to test to 67, and ideally 111, markers. This is the bedrock of your profile and is used in all projects and studies.
Another problem facing testers is the array of tests available which often causes confusion. Projects like this one, which follow patrimonial inheritance, can only use Y-chromosome DNA data. For example, unless you are following matrimonial inheritance for other reasons, testing for mitochondrial DNA is unnecessary. Having Family Finder and/or mt-DNA tests performed will not help you track your paternal ancestry and will be ignored in this Project
Next in importance to STR testing (above) is SNP testing to establish your haplogroup. From your STR results, FTDNA are able to predict your most likely broad haplogroup. This is shown in red and is normally R-M269 for most McMahons and corresponds to the evolutionary stage of early Neolithic Europeans. Big Y, Pack or individual SNP testing is needed to establish further evolutionary branching typically to R-DF13 or R-DF21 for most McMahons (actual test results are shown in green). Many have tested for more recently discovered SNPs which are now being used to define the Project haplogroups.
Another cause for concern among some testers is where no matches are found even when testing to 67 markers. This should not be regarded as a negative result.It is common and simply means that your lineage evolved early (and is outside the FTDNA statistical limits), no other distant relatives have tested or you are the only survivor of your lineage to have made it to the 20thcentury. Sometimes joining other name projects can result in a match.
Over time a fair number of non-McMahon matches have come to light. Some of these are known adopted names while others are suspected of being adopted names or non-parental progeny. Prior to record keeping (mid 19thc.), fostering sons between chieftains was a common practice which no doubt could have accounted for some switching of names and DNA. Before proper surnames were introduced (in and around the second millennium), it is also likely that some tribal/family groups had become genetically mixed before they were known by their surname.