Hidden Branches of the Cumberland Cluster

Since the DYS458.2 cluster bomb in 2013 we’ve entered the world of next gen Y testing like Big Y. That testing has compressed the search for relevant SNPs from a years long trudge down to a matter of a few months (along with a bunch of analysis afterwards). In 2013 we knew we had two branches. With big Y tests rolling in on the DYS458.2 side of the house, we’ve been able to fill in even more branches, down to the family level and sometimes within the family which has been exciting for the Elmers and of course me.

What has been missing for quite a while is a good trail blazed into the territory of the smaller DYS458 (normal) group. Through panel tests, we can compare them to the 458.2 big Y testers, but their own side of the house was an undiscovered country. So for instance Schmidt and Corson/Jansen were negative for ZP85 which has been carried by all the 458.2 men, but it was hard to tell if that was the only SNP that was missing or how it was positioned among them on the tree. Excuse the two Wrights in the graphic…they are the same man. Although having them in the same table illustrates how much you can learn from a big Y test I suppose.

cumberland panel results with big Y

Basically, we know there is at least one branch there that does not have ZP85..we just haven’t exactly found the base and we don’t have any path to follow.

This is where things get a bit strange. Family Tree DNA‘s Big Y test is great, but their matching and results display is…well…hard to follow. A lot of information is missing and some of the information that is included is redundant. The best analysis of the big Y results is actually done by other companies (Yfull and Full Genomes seem to be popular), or by Y DNA projects like the Z18 and U106.

Most of the matches we’ve seen pop up at FTDNA are already part of the Z18 and/or U106 projects, so we expect to see their results anaylized and placed in a logical order by the project admins. Sometimes, that is not the case. Cumberland has a few people who have not submitted their results to the various projects. One of those is a DYS458 (normal) Cumberland listed as “Baker”.  He has blazed a trail through the dark heart of their area, but has not come back to draw a map. So far, shouting into the dark has not gotten anyone a response. Sometimes people wish to remain very private and sometimes they die, it can be hard to tell the two apart.

Needless to say, for the group, that is a big loss. This was my attempt to track his results through an FTDNA big Y “shared novel varient” comparison with one of our Elmer accounts and a comparison to the panel test of Schmidt (458 normal):

comparison of baker and schmidt results

Yeah. My thought is that FTDNA does not think any of the SNPs available in the panel test are “novel”. So the picture was kind of patchy. There is a report of SNPs not shared, but again, I’m not sure which of the known SNPs would be reported there, so it is possible that it would always be patchy. If we combined the two results, we would get a fuller picture, but that also seems dangerous as these men might be thousands of years apart or in different parts of their tree. Baker may not have S4028 or S3524 for instance placing him farther back in time than Schmidt.

It’s just not complete. Here is the visual with the Elmer results in the same area for comparison:

baker comparison to elmer

Lucky for us though our friend Skjennum from Norway (also a normal 458) ran through a big Y this last month and we have his results to compare. Here is a more apples to apples comparison of our Elmer kit to Skjennum:

Skjennum big Y comparison to Elmer

The coverage boundary issues seem to resolve themselves to positives (like ZP84 which is cbl for both tests), but even if we take them out there is a big run of 12 common SNPs including ZP86. I found that surprising. The 458.2 split, based on the panel results and Skjennum, seems to be around the time of ZP85. None of the others appear to be glaringly missing. ZP85 is the last SNP shared with 458.2 positive men, Ovens (Scotland) and Lund (Norway) before they go down their own paths.

To my layman’s eyes it looks like there is a pinch point (a huge die off of other DF95 branches) or population isolation with self selecting for one male line (small isolated village with a high mortality rate for other DF95s), that accounts for our very long chain of common SNPs. If we’re looking at 130 years per SNP as a rough guide, then we’ve got about 1500 years worth of missing branches to account for.

Once we get beyond that pinch point we start to see branches and diversity in our groups. Using the Elmers as a gauge, there are another 14 or so big Y SNPs in their line after the split with ZP85. Another 1500 or 2000 years of doing our own things.  On the ZP85 side we see several new branches and a few that we have a trail for, but no one else has walked down.

On the 458.2 and ZP85 negative side, it looks like some of the men including Schmidt and Baker have a null 425 which puts them in a small group of their own. Corson/Jansen and Skjennum do not have that null 425, hinting at more branches still to be had from their kits as well. I will just mention here that world family names lists Norway as the second most populous “Jansen” country. The Netherlands wins hands down with six times as many Jansens as Norway, but..I’m just putting it out into the universe.

The Z18 project has a nice graphic of the current Z18 tree (sans Z372), this is the portion that contains us:

DF95 and subclades

On the map you can see DF95, really representing the last SNP we all know we have. It’s surrounded by about 13 other SNPs so we don’t know it’s exact position. It may help to think of that little blue block for DF95 as 14 blocks going down in a straight line. Then you hit CTS2207. CTS2207 appeared in the data from the thousand genomes project along with CTS2206 (which is one of the 13 SNPs surrounding DF95. So it appears to be a branch, but no one in our group has been positive for it yet. What you might expect to find under CTS2207 is what we see under ZP85.

ZP85 leads to 458.2. This is where we leave Lund and Ovens who both have big Y tests. What you can imagine stemming from ZP85 are two unknown branches with their own blue boxes. One that contains Lund and his unknown people and the other that contains Ovens and his unknown people. We just don’t know what the major SNPs on their branches are yet. The branch we do know about it ZP121.

ZP121 is my group. You can see ZP124 which is representing a break point where the Lunsfords go off on their own branch. ZP129 represents our branch with the Knowltons before we both split up. On the other branch, ZP125 with Wright, Stanuczek and Winne. Wright goes on his own branch and Winne and Stanuczek continue on to ZP150 before going their separate ways.

Back up to that NUL425 next to ZP85. Like I was saying that group includes our Baker mystery man with unknown branches and blue boxes below it, likely including Schmidt, Chamberlain and our friends from old…the Coens. What you don’t see is the base SNP for Skjennum or Jansen/Corson, which might be contained in the Skjennum results. A matching sibling SNP for ZP85.

For now, we have to imagine these hidden major branches and shared SNPs. We know they are there. We’re just waiting for others to light a torch and move down the trail that was blazed by pioneers.

Here is an updated chart based on the Z18 panel results listed at L257.org along with some comparative big Y results and Skjennum included. I suspect given the Wright and Skjennum big Y results that reporting on S8387 in the panel run is prone to error so I would ignore those negatives until further testing proves them negative.


MORE…always more. We need more men to walk down these paths and blaze trails so that we can have a fuller picture of our various movements and migrations. It would be interesting to see where our German and Benelux testers fall in all this.





Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.