Sample Analysis Using Sortable Spreadsheet

[This is a sample analysis of Y-DNA results using the R1a-Y2619 Ashkenazi Levite Sortable Spreadsheet and the methodology set forth elsewhere on this website, using data concerning my matches as of June 22, 2013. The description of the analytical method remains applicable, but the pool of matches has expanded considerably since the time of this analysis. Brackets in italics indicate comments added on May 21, 2017, based upon the results of additional STR and SNP testing.]

This example uses my Y-DNA markers, which are far enough from the R1a1a Ashkenazi Levite mode for this methodology to provide meaningful results.[1] (At 67 markers, I’m eight steps away from the mode.) As discussed in detail below, my seven closest matches and I appear to share certain characteristic marker values that indicate, I believe, that we share a Most Recent Common Ancestor (“MRCA”) who may have lived perhaps in the early 17th century. Based upon the same analysis, it appears that there is another cluster of six men who share, with my seven closest matches and me, a MRCA who may have lived perhaps in the 16th century. (Alternatively, SNP testing may demonstrate that these two clusters are quite distant from each other, and that their shared deviations from the mode resulted from independent mutations and/or reflect the original R1a1a Ashkenazi Levite mode.)

[Additional results, including the results of Big Y testing, confirm the existence of the two clusters identified in my analysis posted on June 22, 2013, but indicate that the MRCA of those clusters lived considerably longer ago than predicted in that analysis. The two identified clusters both fall within the R1a-FGC18222 cluster of R1a1a Ashkenazi Levites. The MRCA of that cluster likely lived about 1,500 years ago. The men in my cluster (which we now know to be the R1a-FGC18218 cluster) share with me two SNPs downstream from FGC18222 - FGC18226 and FGC18218. The men in the second cluster (which we now know to be the R1a-YP1074 cluster) share two different SNPs downstream from FGC18222 - YP1074 and YP1075. The MRCAs of the R1a-FGC18218 and R1a-YP1074 clusters likely each lived about 1,000 to 1,200 years ago.]

Spreadsheets Used in This Analysis

There are three spreadsheets that I used in this analysis. The spreadsheet that most readily shows the marker values used for the analysis: (1) includes the results for only the 20 men who match me most closely, sorted to the number of steps at which they differ from me; (2) includes the values for only those markers where any of us differs from any of the others; and (3) is limited to deviations on the first 67 markers (when I have more close matches who have tested to 111 markers, I may expand the spreadsheet to include deviations on markers 68 through 111).[2]

The second spreadsheet is the R1a1a Y-DNA Sortable Spreadsheet (v. 7) dated June 1, 2013, sorted to the number of steps at which all men who have tested to at least 67 markers differ from me at 67 markers. (This is the spreadsheet that I used to prepare the first spreadsheet.)

The third spreadsheet is prepared in the same manner as the second spreadsheet, sorted to the matches of a man who is quite close to the R1a1a Ashkenazi Levite mode. It shows that the method described in this spreadsheet will not work well for men who are very close to the mode at 67 markers. Because such men and their closest matches have fewer deviations from the mode, such men are less likely to have patterns of deviations from the mode allowing identification of men who share a relatively recent MRCA with them. For such men, upgrading to 111 markers will generally be necessary in order to identify and confirm close matches; men who are at or near the mode at 111 markers may still be unable to identify close matches unless they share deviations from the mode on at least two markers (at least one of which would be, preferably, slow mutating).

Sorting the Spreadsheet by Your Closest Matches

I started this process by inputting, into the columns headed “Steps at 67” and “Steps at 111,” my genetic distance from my closest matches.[3]

For men with numerous matches at 111 markers, it may be useful to sort the spreadsheet to 111 markers. Because I have so few matches at 111 markers even going out beyond the 10 markers reported as matches by FTDNA, sorting the spreadsheet to 67 markers works best for me.

After inputting my genetic distance at 67 markers from men who have tested to at least 67 markers, I sorted the spreadsheet by right-clicking on the “Steps at 67” column in the top row with a man’s name and marker information, clicking “Sort,” and then clicking “Sort Smallest to Largest.”

For purposes of illustration, I then pared down the spreadsheet: (1) to include only those men with a genetic distance of eight markers or less; and (2) to eliminate all of the marker values on which neither I nor any of my matches have a deviation from the R1a1a Ashkenazi Levite mode.

Analyzing Patterns on the Spreadsheet

Although individual mutations will occur randomly and may often occur independently, it is less likely that the same mutations occurred independently among a group of men who share multiple deviations from the mode, especially to the extent that the deviations appear on markers that mutate slowly. (Conversely, the fact that men share deviations from the mode on more than one fast-mutating marker may be consistent with independent mutations, especially if the men do not share deviations on slow-mutating markers.)

As the spreadsheet shows, I have eight deviations from the R1a1a Ashkenazi Levite mode on the first 67 markers. Seven of those deviations are on fast-mutating markers, so they could reflect relatively recent mutations – which would be of little genealogical significance, except for very close matches – or more remote mutations – which would be more likely to be of use in identifying men who share a relatively recent MRCA.

My single deviation from the mode on a slow-mutating marker is on DYS537, where I have DYS537=11, as compared to the mode of DYS537=12. Most of my closest matches share DYS537=11, so that marker value appears to be of genealogical significance. Furthermore, I share DYS537=11 with more of my closer matches than I share any other deviation from the mode, which makes it more likely that this was the first deviation from the mode that occurred on this line.[4]

There are three markers where I share a deviation from the mode with six or seven of my closest matches – DYS607=17 (rather than the mode of DYS607=16), DYS570=21 (rather than the mode of DYS570=20), and DYS446=13 (rather than the mode of DYS446=12). Although these are fast-mutating markers, the fact that I share all three marker values with my closest matches makes it highly likely that the mutations were inherited from a shared MRCA.

I share another marker value, DYS460=12 (rather than the mode of DYS460=11) with my three closest matches. This marker value may reflect a more recent branching of lines among my closest matches.

I share DYS576=20 (rather than the mode of DYS576=19) with my closest match, but not with my four next closest matches. The match on DYS576=20 with my closest match presumably reflects a mutation inherited from our MRCA. The fact that I share DYS576=20 with some of my other close matches (but not my closest matches) probably reflects independent mutations, but there is a possibility of a back-mutation among my four next closest matches.

Finally, I have two deviations from the mode (DYS458=15 and DYS449=31) which are not shared by my closest matches. Those deviations are likely to be recent, so they are likely not of genealogical significance (except for identifying very close matches). Accordingly: (1) my variances from other men on these markers are entitled to little weight in evaluating relationships; and (2) my matches with other men on these markers are likely to reflect independent mutations, and the matches therefore appear to be closer than they likely are.

Analyzing Relationships with Closest Matches on a One-by-One Basis

As discussed in detail below, my seven closest matches and I appear to share an MRCA who had four deviations from the R1a1a Ashkenazi Levite mode: (1) DYS607=17; (2) DYS570=21; (3) DYS537=11; and (4) DYS446=13. Of these markers, DYS537 is the most significant, because it is the slowest mutating marker. The marker values of these eight men form a neat pattern that can be analyzed to determine the order in which mutations likely occurred.

Czerwick (220657). I do not have any known relatives on my direct Wexler line who have done Y-DNA testing. However, my closest match at 67 markers (64/67), Czerwick, is descended on his direct male line from a man, Franciszek Czerwik, who lived in Bargly, Poland, 10 miles south of Czestochowa, Poland, as of the early 18th century. By 1809, my Wexler ancestor, Salamon Wexler, was living in Klobuck, Poland, 10 miles north of Czestochowa. Given this geographic and genetic proximity, it appears likely that Czerwick and I share an MRCA who lived sometime in the 18th century; our FTDNA TiP report predicts an 88.67% likelihood that we share an MRCA within 12 generations, and a 97.08% likelihood that we share an MRCA within 16 generations.[5]

By comparing the three markers on which Czerwick and I differ with the marker values of my closest matches, I can confirm that the marker values that we do not share reflect relatively recent mutations. On two of the three markers where we differ, I have deviations from the mode (DYS458=15 and DYS449=31) that I do not share with any of the men who match me within six markers or less; on the third marker where we differ, Czerwick has a deviation from the mode (CDYa=34) that is not shared by any man within seven markers of me. Thus, our deviations from the mode on those three markers likely represent mutations that have occurred on our individual lines since the time when our MRCA lived.[6]

[Subsequent a la carte SNP testing has confirmed that Czerwick and I share three SNPs downstream from the FGC18215 level not shared with our other matches who have done Big Y testing (FGC18221, FGC18223, and FGC18225), and that Czerwick is therefore my closest match among the R1a-Y2619 Ashkenazi Levites who are discussed in this analysis.]

Accordingly, the MRCA that Czerwick and I share likely had the six mutations from R1a1a Ashkenazi Levite mode that the two of us share: (1) DYS460=12; (2) DYS607=17; (3) DYS576=20; (4) DYS570=21; (5) DYS537=11; and (6) DYS446=13.

Gann (291), Bliss (208320), and Mordecai G. (privately provided). My next closest matches (63/67) are Gann and Bliss. They are 65/67 matches for each other. They both may share a direct male ancestor from Lithuania; Bliss’ direct male ancestor was named Jaffe, and there is a Gann family tradition of a relationship with a family named Jaffe. Because Gann has traced his ancestry back to the early 19th century, his MRCA with Bliss presumably lived sometime in the mid- to late 18th century. Mordecai G. is a known distant cousin of Gann, sharing an ancestor in the early 19th century, but he is only a 64/67 match for Gann.

Gann, Bliss, and Mordecai G. differ from the hypothetical MRCA whom Czerwick and I share on DYS576, where they all have the mode of DYS576=19 (while Czerwick and I have DYS576=20).

On DYS460, Gann shares DYS460=12 with Czerwick and me, while Jaffe and Mordecai G. have the mode of DYS460=11.

On DYS444, Gann shares the mode of DYS444=15 with Czerwick and me, while Jaffe and Mordecai G. have DYS444=14.

Based upon these marker values, I assume that: (1) the MRCA that I share with Czerwick and Gann likely had the four mutations from the R1a1a Ashkenazi Levite mode that the three of us share: (a) DYS607=17; (b) DYS570=21; (c) DYS537=11; and (d) DYS446=13; and (2) Gann’s line, after breaking off from Mordecai G.’s line, had a mutation to DYS460=12 and a back mutation to DYS444=15.

[Subsequent Big Y testing has confirmed that Gann, like Czerwick and me, is FGC18222+ FGC18226+ FGC18218+ and that the three of us also share three SNPs downstream from FGC18218 (FGC18215, FGC18220, and FGC18227).]

Accordingly, the MRCA that Czerwick, Gann, Bliss, Mordecai G., and I share likely had the four mutations from the R1a1a Ashkenazi Levite mode that the five of us share: (1) DYS607=17; (2) DYS570=21; (3) DYS537=11; and (4) DYS446=13.

Phillips (40651). My next closest match, Phillips (62/67), is descended from a line that lived in London by the late 18th century. There are three markers on which he differs from the assumed Wexler/Czerwick MRCA. First, he has DYS393=12; because that marker value is not shared by any of our closest matches, it presumably reflects a relatively recent mutation in his line. Second, he has the mode of DYS460=11, while the assumed Wexler/Czerwick MRCA had DYS460=12; this presumably indicates that the mutation DYS460=12 occurred after the split between the Phillips line and the Wexler/Czerwick line. Third, he has the mode of DYS576=19 like the Gann/Bliss line, but unlike the assumed Wexler/Czerwick MRCA, who had DYS576=20; this tends to confirm that the original marker value of the shared ancestor was DYS576=19 (which, not surprisingly, is the mode).

[Subsequent Big Y testing has confirmed that Phillips, like Gann, Czerwick, and me, is FGC18222+ FGC18226+ FGC18218+ and that we also share three SNPs downstream from FGC18218 (FGC18215, FGC18220, and FGC18227). In other words, Phillips, Gann, Czerwick, and I share an MRCA.]

Accordingly, the MRCA that Czerwick, Gann, Bliss, Mordecai G., Phillips and I share likely had the four mutations from the R1a1a Ashkenazi Levite mode that the six of us share: (1) DYS607=17; (2) DYS570=21; (3) DYS537=11; and (4) DYS446=13.

Kowalsky (175107). My next closest match is Kowalsky (61/67). He differs from the assumed Wexler/Czerwick MRCA on: (1) DYS460, where he has DYS460=10 (one step below the mode), the Bliss/Mordecai G./Phillips (and, presumably, Gann) line has the mode of DYS460=11, and the assumed Wexler/Czerwick MRCA has DYS460=12 (one step above the mode); (2) DYS576, where he has the mode of DYS576=19 (like the assumed Gann/Bliss/Phillips MRCA), as compared to the DYS576=20 of the assumed Wexler/ Czerwick MRCA; and (3) DYS557, where he has DYS557=16 (unlike all of our other closest matches, who have DYS557=15, indicating that this is a recent mutation in the line).

[Subsequent Big Y testing shows that Kowalsky belongs to the same cluster as Gann.]

This analysis shows that a marker is equally likely to mutate to a larger number of STRs or to a smaller number of STRs, and that a two-step deviation may mask a relatively close relationship. Accordingly, the Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky MRCA likely had the four mutations from the R1a1a Ashkenazi Levite mode that the seven of us share: (1) DYS607=17; (2) DYS570=21; (3) DYS537=11; and (4) DYS446=13.[7]

Schiff (N107659). My next closest match is Schiff (61/67). He differs from the assumed Wexler/Czerwick MRCA on: (1) DYS449=29 (one below the mode, as compared to my DYS449=31, one above the mode, and Czerwick/Gann/Bliss’ DYS449=30, at the mode); (2) DYS570, where he has the mode of DYS570=20 (as compared to the Wexler/Czerwick/ Gann/Bliss/Mordecai G/Kowalsky/Phillips DYS570=21); and (3) CDYb=39 (unlike all but one of our closest matches, who have the mode of CDYb=38). He also has DYS576=20, which matches Wexler/Czerwick, but not Gann/Bliss/Mordecai G./Kowalsky/Phillips.

Schiff and I have a two-step deviation on DYS449 that makes our relationship seem less close than it is. Because our other closest matches have DYS570=21 and because DYS570 mutates quite quickly (relatively speaking), my guess is that Schiff’s DYS570=20 is a back mutation from DYS570=21. Because Gann/Bliss/Mordecai G./Kowalsky/Phillips have the mode of DYS576=19 and because DYS576 also mutates quite quickly, my assumption is that Schiff’s DYS576=20 matches Wexler/Czerwick’s DYS576=20 as a result of an independent mutation.

[Subsequent Big Y testing has shown that Schiff, like Phillips, Gann, Kowalsky, Czerwick, and me, is FGC18222+ FGC18226+ FGC18218+ but that, unlike the other men, Schiff does not share any of the three SNPs downstream from FGC18218 (FGC18215, FGC18220, and FGC18227).]

I therefore conclude that the Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky/Schiff MRCA likely had the four mutations from R1a1a Ashkenazi Levite mode that the eight of us share: (1) DYS607=17; (2) DYS570=21; (3) DYS537=11; and (4) DYS446=13.

Analyzing Relationships with Somewhat More Distant Matches on a One-by-One Basis

My 13 remaining matches within eight markers or less do not share DYS446=13 with my eight closest matches and me.

As discussed above, the most distinctive marker value that I share with my closest matches is DYS537=11 because that marker is a slow mutating marker; there is some chance of back mutations on that marker, but the chance of a back mutation on that marker is considerably less than the chance of a back mutation (or an independent mutation) on other markers.

Based upon that assumption, the eight men among my 13 remaining matches within eight markers or less with DYS537=11 are more likely to share a more recent MRCA with Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky/Schiff than are the five men among those matches with DYS537=12.[8]

Six of the eight men with DYS537=11 also have DYS570=21, the assumed marker value of the Wexler/ Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky/Schiff MRCA. However, six of these eight men also have DYS459b=11, while Wexler/Czerwick/Gann/Bliss/ Mordecai G./Phillips/Kowalsky/Schiff have the mode of DYS459b=10. DYS459b is a medium-mutating marker, which makes independent mutations less likely.

My assumption is that these six men – Graff (107789), Muller (227160), Christmann (N103033), Howards (21822), Sacks (213421), and Gordon (139616) – may have shared an MRCA who had DYS459b=11, DYS570=21, and DYS537=11.[9]

That would mean that Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky/Schiff/Graff/Muller/ Christmann/Howards/Sacks/Gordon may have shared an MRCA who had: (1) DYS570=21; and (2) DYS537=11.

Under this hypothesis, the Graff/Muller/Christmann/Howards/Sacks/Gordon line thereafter had a mutation of DYS459b=11.

Note that this hypothesis is equally consistent with: (1) the two lines having independent mutations on one or both of DYS570 and DYS537; or (2) the two lines had a shared MRCA with DYS570=21 and DYS537=11. As discussed elsewhere on this website, it appears that the R1a1a Ashkenazi Levite progenitor likely had DYS537=11 and that the MRCA of the modal line had a mutation to DYS537=12. If so, the MRCA of Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky/Schiff and Graff/Muller/Christmann/Howards/Sacks/ Gordon could well have pre-dated the MRCA that one of those two branches shares with the Horowitzes. SNP testing should confirm, or refute, this possibility.

[Subsequent Big Y testing of Graff has shown that he is R1a-FGC18222 but that he belongs to a different branch than Czerwick, Gann, Kowalsky, Phillips, and me - the R1a-YP1074 branch. None of the other men identified in my 2013 analysis had done Big Y testing as of May 2017. Between May 2017 and July 2021, Muller and Christmann received Big Y results that showed that they, like Graff, belong to the R1a-YP1074 branch.]

Confirmation by Considering Results at 111 Markers

Ideally, I would be able to confirm – or refute – the validity of my analysis at 67 markers by comparing results at 111 markers. Three of my seven closest matches at 67 markers - Mordecai G., Phillips, and Schiff - have upgraded to 111 markers; their marker values continued to be close to my marker values, tending to support the analysis set forth herein.

For purposes of this discussion, the most significant marker on markers 68 through 111 is DYS650. The mode is currently DYS650=20; because the results are split and there are some men with DYS650<19 and no men with DYS650>20, I believe that the original mode was DYS650=19.

I have DYS650=18, two steps from the mode. My hypothesized closest matches among the men who have tested to 111 markers, Mordecai G., Phillips, and Schiff, have DYS650=17 – a marker value that indicates that their lines at one time had DYS650=18, an intermediate step from the mode (whether it be DYS650=20 or DYS650=19) to DYS650=17. [Subsequent testing has shown that Czerwick shares DYS650=18, confirming his status as my closest match who has tested to at least 67 markers.]

Muller, Howards, and Christmann all have DYS650=19, which is consistent with my hypothesis that those three men share an MRCA who lived more recently than the MRCA whom I believe that I share with my closest matches, including Mordecai G., Phillips, and Schiff. (If DYS650=19 was the mode of the R1a1a Ashkenazi Levite progenitor, the fact that Wexler/Mordecai G./Phillips/Schiff share DYS650<20 with Muller/Howards/Christmann may indicate that they are descended from the R1a1a Ashkenazi Levite progenitor, rather than that they share an MRCA who post-dates their branches' split from the branch that is close to the R1a1a Ashkenazi Levite mode.

Muller and Christmann – but not Howards – share DYS513=13 (as compared to the mode of DYS513=12), which suggests that their shared MRCA lived more recently than the MRCA that they share with Howards.

Garbarz has DYS650=20, the mode, which is consistent with my belief that my apparent close match with him reflects distance to the mode, rather than a close relationship as evidenced by specific mutations.

Tying the Spreadsheet Analysis to Predicted Time to MRCA

To evaluate the accuracy of my analysis of marker values set forth above and to try to determine the appropriate time frame, I compiled information from two sources of estimates concerning the amount of time to an MRCA between my closest matches and me.[10]

First, I considered FTDNA’s TiP reports and used them to determine the number of generations to an MRCA with a 50%, 75%, and 95% degree of certainty (i.e., that there is a 50%, 75%, or 95% likelihood that the time to an MRCA is equal to or less than the number of generations at which such percentage is reported). Unfortunately, those reports provide estimated time to an MRCA at 67 steps only for men who are at a genetic distance of seven markers or less to me.

Second, I ran the Y-Utility function to calculate the number of generations to an MRCA with a 50%, 75%, and 95% degree of certainty.

The following table identifies my closest matches, in the order in which I grouped them above, and shows the calculated time to an MRCA using the TiP reports and Y Utility.[11]

The gradation in time to an MRCA is generally consistent with the order of mutations that I hypothesized in my analysis, above. Because my hypothesized order of mutations based upon shared deviations is, for the most part, inversely related to the number of steps in genetic distance, however, it’s unclear whether the TiP report and the Y-Utility results support my analysis or merely reflect the correlation between my analysis and decreased genetic distance.

Because the men who are at a genetic distance of 7 or 8 from me but do not share a pattern of distinctive markers with my other closest matches are predicted to have an MRCA with me who lived at about the same time as the MRCA shared by my closest matches and me, I suspect that that the TiP report and the Y-Utility results do not confirm my analysis.

The time periods calculated by the TiP report and the Y-Utility results are, at a 50% degree of certainty, generally consistent with the timeline that I believe to be likely with regard to my closest matches. In particular: (1) my MRCA with Czerwick is predicted to have lived in the early 18th century, which makes sense, given that (a) we are 64/67 matches whose mutations appear to be relatively recent and (b) our direct male ancestors lived near each other but were Catholic and Jewish, respectively, as of the early 19th century; (2) the Wexler/Czerwick MRCA with Gann, Bliss, and Mordecai G. is predicted to have lived in the latter part of the 17th century; and (3) the Wexler/Czerwick/Gann/Bliss/Mordecai G. MRCA with Phillips, Kowalsky, and Schiff is predicted to have lived in the early 17th century.

The Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/Kowalsky/Schiff MRCA with Graff, Muller, Christmann, Howards, Sacks, and Gordon is predicted to have lived in the 16th century. If, as discussed above, the shared deviations between these branches were inherited from the R1a1a Ashkenazi Levite progenitor, the actual time to an MRCA would be considerably longer than the calculated amount; as discussed above, SNP testing should provide insight into whether these STR-based calculations overstate the relationship between these two clusters.

[Based upon the Big Y results, the predicted times to an MRCA discussed above seem considerably too short (with the possible exception of the predicted time to the MRCA that I share with Czerwick). In particular, of the men who have done Big Y testing: (1) the R1a-FGC18215 ancestor shared by Wexler/Czerwick/Gann/Phillips may have lived 600 to 800 years ago; (2) the R1a-FGC18218 ancestor shared by Wexler/Czerwick/Gann/Phillips/Schiff may have lived a century or two before the R1a-FGC18215 ancestor; and (3) the R1a-FGC18222 ancestor shared by Graff with Wexler/Czerwick/Gann/Phillips/Schiff could have lived 1,000 to 1,200 years ago or more.]

-----------------------

[1] Unfortunately, this methodology does not work very well – or, perhaps, not at all – for men who are close to the mode, since they are less likely to share a pattern of mutations with their closest matches. I’ve tried to perform the same analysis for a man, Vodovoz (248865), who is very close to the mode (within two steps at 67 markers and within four steps at 111 markers). As review of the spreadsheet will demonstrate, it’s virtually impossible to find patterns of matches at 67 markers (especially since one of his mutations on the first 67 markers is unique and therefore of no genealogical significance, except for very close matches), but certain patterns may be discernible at 111 markers.

[2] The spreadsheet also includes information concerning an estimated time to an MRCA in generations compiled from three sources: (1) the FTDNA TiP reports (I ran the reports on a generation-by-generation basis, and I used the number of generations to an MRCA that was closest to the TiP report’s estimates at 50%, 75%, and 95%); (2) the Y-Utility function (instructions on using that function are posted in the instructions posted elsewhere on the group website); and (3) a Janzen spreadsheet (the posted results appear to understate the time to an MRCA, presumably because the matches are so close; I haven’t given those results any weight in my analysis).

[3] You can find this information by using the sortable spreadsheet posted on this website, in conjunction with the steps to matches spreadsheet posted on this website, following the steps explained on this website. You can also find this information using your FTDNA match list. Finally, you can find this information by using the Y-DNA search utility on the [now defunct] Semargl website, which allows you to generate a table showing your genetic distance (by number of markers or by total alleles) from your matches. The Semargl utility allows you to generate a table for your matches going out to 15 markers or more; by contrast, the FTDNA website shows matches up to a genetic distance of only seven at 67 markers. Because the spreadsheet includes marker values for some men whose results are not reported on the Semargl website, some genetic distances will need to be counted manually.

[4] I note, more generally, that DYS537=11 is almost as common as DYS537=12 among the R1a1a Ashkenazi Levites for whom I have search results (the split is currently about 45% to 55%). The known descendants of the Horowitz rabbinical family have DYS537=12, as do the men surnamed Horowitz for whom we have test results at 67 markers; the Horowitz men with known genealogies descend from one line of the Horowitz family. I believe that this deviation likely reflects a relatively early split in the R1a1a Ashkenazi Levite lines. Because the number of men with these two marker values is so evenly split, it’s not clear whether the original value was DYS537=11 or DYS537=12; I suspect that the original value was DYS537=11 because there is considerably more deviation in marker values among men with DYS537=11 than among men with DYS537=12, which suggests that the cluster of men with DYS537=11 is older than the cluster of men with DYS537=12.

[5] Interestingly, in the early 1800s, the Czerwik line was Catholic, while the Wexler line was Jewish. I speculate that the Czerwik line may have been Frankists – followers of Jakob Frank, an apostate who converted to Catholicism; he was imprisoned in a monastery in Czestochowa in the 1760s, and many of his followers – who also converted to Catholicism – moved to the Czestochowa area.

[6] While there is theoretically a chance that our MRCA had the deviation in the marker value and that there was a back mutation in one of our lines, the odds of a back mutation are low because our closest matches do not share the deviation from the mode.

[7] These are the same four mutations shared by the Wexler/Czerwick/Gann/Bliss/ Mordecai G./Phillips/Kowalsky MRCA. Put otherwise, Kowalsky and Phillips may, based upon their marker values, share the same MRCA with Wexler/Czerwick/Gann/Bliss/ Mordecai G.

[8] Three of my five matches with DYS537=12 – no name identified (110775), Garbarz (190924), and Cornbleet (95663) – are within two markers of the R1a1a Ashkenazi Levite mode. They share only one mutation (or, in one instance, no mutations) with Wexler/Czerwick/Gann/Bliss/Mordecai G./Phillips/ Kowalsky/Schiff, so their relatively close genetic distance presumably reflects the fact that they are at or near the mode; it is not possible to tell whether a close match to someone at the mode reflects a close relationship or a lack of mutations over a longer period of time.

I suspect that the fourth of these five matches – Segil (N55713) – appears to be a closer match to me than he is because of an independent mutation in his line of DYS458=15. The fifth of these five matches – Greenberg (63501) – is discussed below.

[9] It is conceivable that Howards (21282) could be part of the Wexler/Czerwick/Gann/ Bliss/Mordecai G./Phillips/Kowalsky/Schiff group, given that he shares with their presumed MRCA: (1) DYS607=17; and (2) DYS570=21. For this to be the case, his line would have needed to have a mutation on DYS459b, from DYS459b=10 to DYS459b=11, that was independent from that mutation on the Graff/Muller/Christmann/Howards/Sacks/ Gordon line. Because DYS459b mutates at a medium rate, this scenario is less likely than the scenario hypothesized above. Greenberg (63501) may be part of this group, because he has DYS459b=11; to belong to the group, however, he would need to have had a back-mutation on DYS537, from DYS537=11 to DYS537=12. The pattern may be clearer at 111 markers.

[10] I also tried using a Janzen spreadsheet to calculate the amount of time to an MRCA for the men who match me at genetic distances of 5, 6, 7, and so on. Those results do not appear to be of much use to this analysis, given that my matches at a genetic distance, of e.g., seven steps may not be closely related to each other.

[11] Ben Weinstock has observed that the time when the TiP report predictions first increase by less than 5% per generation matches closely with known genealogical relationships of tested men. Except for men who are very closely related, this 5% decrease typically occurs within a generation or two of the generation at which the TiP report is at 75%.