Ornery STR Matching (OSM)

We are looking for better ways to assist those who have Ornery STR Matching (or OSM for short). (Note: we have created a new term.) We have to profusely apologize to some new members when we see their test result and realize they are a member of the Ornery STR Matching group. And again when we ask them to upgrade to 67 or even 111 STR markers and the problem persists. This study is to document and discover if we can do more to alleviate the pain of what seems like a crap shoot when it comes to STR testing and finding Patrilineal relatives for some lines.

OSM is somewhat a by-product of how or which STR's happened to be chosen in the first 12, 25, 37 and so on test panels at FamilyTreeDNA. The few other yDNA STR companies likely exhibit the same issue but we do not have enough samples to verify. The issue arises, we believe, due to the fact that a small sampling of the available yDNA STR's is being used. The expansion of Sequencing testing like BigY and its use to extract values for many of the 500+ STR values may level the playing field for all if that becomes the new first approach to yDNA STR testing. But, until then, the issue described here will persist. The problem presented here is in contrast to some other study members who, with just a y12 test, can verify members in their family branch (that can then be further confirmed by deeper STR and deep SNP testing). Furthermore, only members of their family branch are in the match group. For these members, it appears, they just so happen to have a few very unique STR marker values (among the general population) that no one else has (at all, or in combination). Their match lists, at pretty much all levels, only contain other Patrilineal family members. This as compared to OSM members who have y37, 67 and 111 match lists with 10's to 100's of entries; none of which seem a Patrilineal relative. And often with different clades when (deep) SNPs are tested.

We hypothesize the root of the problem is which STR values are being tested. That the current set in y12, y25, y37 and so on is good for some lines at isolating and differentiating them from the general population. But very poor for others. And that once one gets to all 500+ values, the GD of matching members and non-members of a Patriline will be the appropriately given for all — whether OSM or not, and match lists will indicate Patriline relatives only. Is there a "right" y12 panel for all haplogroups that differentiates them better? Much like the current panel does for one family branch here? Can a better y67 panel be developed that covers more differentiation for a wider swatch of the population? These are lofty questions that cannot be studied with our limited access to data here. So, in the mean time, is there some better advice we can provide OSM members other than trying to do BigY with 500+ STR marker extraction and then possibly wait for potential matches to do the same?

Let us first characterize what we mean and then how we plan to develop a study to work on solutions further.

Characteristics of OSM

Characteristics of those who belong to the Ornery STR Matching (OSM) group:

• y37, 67 and 111 match lists are filled with many, seemingly unrelated surnames and matches; most at the distant extreme edge (high GD for the level of testing).
• Those in the match list with deep SNP clade testing show different haplogroups

Purpose of the Study

To characterize the members to see which specific clades exhibit this issue.

• Are they clades earlier-in-time in the tree and all haplogroups below exhibit the problem? Or are they just unique, deeper branches where those above do not exhibit the issues? That is, if we get a particular clade identified for someone (via STR predicted, SNP Pack or similar means), can we apply techniques developed here to help the member along?
• Are there techniques other than BigY and full 500+ STR value extraction to assist members of such identified clades
• Is the mythical Convergence the source of the problem? If so, why do these reconverge and not others?

Approach to Study

• Identify members with the problem
  • Look at their match lists for various levels and see if discordance of haplogroups (not predicted) exist
  • Determine close and distant GD members of the match list and characterize which are likely true matches (possible?)
• Identify non-members
  • Do their match lists have (m)any non-members?
  • Is there a discordance of haplogroups?
• Look at above results
  • Any patterns emerge?

Example 1

This first example we would never have identified as OSM because we started straight with y37 where he initially had no matches. He has since upgraded to y111 and BIgY. But going back and looking at the FTDNA advanced match table, we see he exhibits the issue in y12 and y25. He has 379 y12 matches and 175 y25. But a single y37 and a different, single y67 match, and NO y111 matches. His one y67 match did y111 but does not match at lower or higher levels. He did not do BigY but shows up in the match list with haplogroup R1a-YP1361. Which is one step above this example testers haplogroup of R1a-YP1363. The example tester has 12 matches in his haplogroup on the BigY match list but all with 8 or more non-matching variants. (It is clear there are numerous testers there that have clumped matching variants that need to be pushed down another branch in the tree.) This tester is not in this project but a good example of how quickly large match lists can drop off and still have no / minimal matching farther out. Even though they have no close y111 STR matching, they still have good "just outside the genealogical time frame" matching in BigY; as of now.

Items of note:

• Of the 379 y12 matches, only 61 did not test higher to explain not appearing on the y25 match list. Of the 175 y25 matches, only 4 did not test to y37 or higher.
• 76 of the y12 matches continued onto the y25 match list. So 99 of the y25 matches are newly introduced as they were not ((y12) matches. 242 of the y12 matches dropped off at y25 due to simply no longer matching.
• Only 1 of the y25 matches continue onto the y37 match list. Only 4 did not test higher so 175 dropped off the match list at y37 due to no longer matching
• 25 matches were GD 0 at y12 and y25. All those tested to at least y37 and are not matches there. The single y37 match is on the earlier match lists as GD 0 and 1; respectively. The y67 match is newly introduced at that level.
• 19 of the y12 and y25 matches tested to y111. 8 who appeared only on the y25 tested to y111. 79 who appeared on the y12 match list only tested to y111

Example 2

Ref: FTDNA Big Y * YSEQ * YFULL * FGC - NGS Discussion Forum Facebook Group post (private group)
58 matches at y111. A number did BigY and all cluster around R-BY3368 and below. (P312 - DF13 - DF5). Young haplogroup branch area. Few novel variants between branches.

OSM in this Project

To help understand whether and how this is a problem in this study, we list in the table below the various branches in this surname study and at what level of testing their match list consists only of genealogical time frame matches. That is, only matches to the surname or maybe known or discovered NPEs as well. There is further detail available for project administrators.

Branch	Test Level	Haplogroup1	Notes
B04	y111	R1b-Z30496 (DF19)	No y111 matches yet 39 y67 distant matches (13 with BigY, 17 only tested to y67); B04DNA, yFull Y17483 3 BigY matches but 23+ non-matching variants each; 12 BigY matches one hapllogroup up, No common surnames among any in BigY match list
B03	y111+	R1b-?	Only tested to y67. 5 GD 5+ matches at y67; only 1 tested to y111. 3 do have common surname with different EKA. No SNP testing to predict haplogroup. 139 y12 matches, 215 y25 matches, 3 y37 matches (no similar surnames). y37 and y67 single common match is only that tested to y111.
B23	y111+	I1-S1990 (pred)	Only tested to y67. 87 y67 matches at GD 5+; no apparent common surnames. 149 matches at y37 and 1,157 matches at y25. Only 2 BigY testers in that. Both under I-S1990 (yFull Y4015 equivalent)
Bxx (KH)	y111	R1b-Z8 (pred) (U106 - L48)	Only tested to y67. 14 of 22 matches in y67 only tested to y67, all GD of 7; few common surnames. 2 BigY tested and under R1b-Z8 (yFull Z8)
B09 et al	y37	R1b-BY4179 (U152 - Z192)	Not OSM! Large y37 match list with many surnames; but all NPE's from one branch. (yFull Y29879)
B10 et al	y12	R1b-Y22958 (L20)	Not OSM! 2 very unique values in y12; one a GD of 2 from most others that likely changed by 2
B00

¹ The link in the Haplogroup is to FTDNA's public tree

External References

• Genetic Distance in relation to level of testing at FTDNA