DF95 – All Together Alone

One of the really striking things about Y DF95 or the Cumberland Cluster, is that we generally tend to have fewer Y matches than others in R1b and R1b-U106 and I imagine even other men in R1b-Z18. In my experience, using Y STRs, we have less false positive matches at lower levels than other haplogroups in R1b. This is because we all have some weird values at places like DYS385a and b (11-11 vs. the standard 11-14 for R1b). We have a really clear shared pattern of Y STRs that is different from the majority of testers, which you would assume makes us fairly closely related. Yet we’re a really old group, apparently branching right off of R1b-Z18. Old and young at the same time.

In Hidden Branches of the Cumberland Cluster I mentioned the cluster of about 14 SNPs right at the top of all of our big Y results. Making our signature SNP, DF95, really 1/14th of the story. In that big long list of shared SNPs are clues to the story of why we are so old, but so similar and it’s a story about pinch points.

As results have come in, I’ve been waiting to see one big Y test or another break up that gang of 14, the way tests have broken up the gangs below it. It just hasn’t happened yet. I’ve also been waiting around for a connection to some other group under Z18, thinking we would tie back into one of them, but all the group analysis shows us branching directly from Z18.

Through the Eye of the Needle

SNPs rack up over time on lines of descent. Instead of a family tree, imagine a descendant tree with the trunk on top and branches going down.

let’s say I make an SNP to add to those that I inherited. My son may be a duplicate of me on the Y, his son may be a duplicate, his son may be a duplicate, but maybe four sons down the line, something changes and a new SNP is born on our line. That 2nd great grandson will have a Y SNP that he will pass on to future generations that is slightly different from mine. My other 2nd great grandsons won’t have that particular SNP, but they may have some others of their own. These form Y branches from my original Y DNA. All of them carry my original set of Y SNPs (that I inherited from my male ancestors) and converge on me as the most recent common ancestor.

The thing is, even though they all converge on me, I’m not the only one who carries these ancestral SNPs, there are many living men that do, even though they may be farther back on the tree. My first cousin probably has an identical set of SNPs, my third cousin, may have a few of his own along with carrying the SNPs we inherited from our shared 2nd great grandfather.

if all the male lines of my descendants save one died out, they could still compare their Y to my 2nd cousin’s descendants and come up with the shared SNPs from my 2nd great grandfather. They would completely bypass me and their last common ancestor and end up at my second great grandfather. My unique SNP would show up in a jumble with theirs on this single living line of descendants.

My second cousin could have twelve sons and their Y line would dominate the picture of our family. My Y line, shrunk to just one man, may die out. My second great grandfather’s Y DNA would continue through my third cousin’s descendants, but my Y SNPs would be on a dead end.

This sort of thing probably happens all the time, with some branches dead ending and others expanding. Below is an example, the different colors represent the formation of new Y SNPs. Any one line may die out but there are others there who carry on. Each little man here is a generation and SNPs might pop up every 140 years or so (judging by big Y).

descendants tree with some lines dead ending

When looking at that list of 14 SNPs shared by every DF95 man, but not shared by anyone else, you get a different sort of picture. 14 SNPs in a row times about 140 years is roughly 2000 years. So every DF95 man alive today has 2000 years of SNPs leading up to our last common ancestor and we don’t share them with anyone else. That means at a certain point 2000 years worth of descendants disappeared, leaving only one man.

cumberland men missing

I don’t have the space to make a 2000 year descendant chart so You’ll have to visualize with the chart above. In the chart above you have the Y line coming from one of those original 14 SNPs. On the right hand side is the DF95 Cumberland line. No one knows what order those 14 SNPs go in, they are a jumble. So I just picked. As you trace the cascade of men down you can see that there would be a lot of inter-related lines and groups that would share some of those 14 SNPs along the way.

We don’t have that. We’re related to that one man at the end. ZP86. All Cumberlands have it along with the rest of the 14. Beyond ZP86 we begin to diverge again forming the branches in the Cumberland Cluster.

That is why we all look so similar and so different from other Z18 men. There should be a lot more of us, maybe some with a few of the differences we have, but not others. In the context of the family tree that should exist, we DF95 Cumberlands would not be so striking. Right now, that man, Mr. ZP86 is our lifeline. He passed through the eye of the needle. Everyone else for 2000 years, is gone.

Broken Branches

Another way to visualize this is by thinking about the branches that should exist. I’m going to put up a few examples so you get the idea of what a Y family tree might look like (again it’s upside down because it’s a descendant tree).

Here is a top level tree for all of R1b made by Mike Walsh from the R1b and all Subclades project. See the tiny DF95 in purple on the right. On the left are some age estimates for major branches. We’re under P311 and U106.

R1b Chart

the tree above is big and messy. Each of those SNPs in boxes may have many many SNPs below them as children, this is just the upper level. You can see that U106 is smaller than it’s sibling P312. Z18 is smaller than it’s sibling Z381 (which is the major group under U106). Still Mr. R1b has a lot of living descendants.

Here is a visual tree made by Iain McDonald of the U106 group. Notice the layering of SNPs on the right where descendants were plentiful and new SNPs seem to crop up fairly routinely in various generations. Note Z18 on the left, with Z17 having child SNPs in Z372 and L257 and the long straight line farther left that leads to DF95 (among others). Click on it for a bigger version…although still pretty tiny to my eyes.

u106 visual tree

From age estimates also provided by Iain McDonald, this graphic above represents R1b-U106 from about 2900BC to today. Very roughly, about 5000 years of history. Lots of children, with some branches thriving and others surviving.

Here is an example closer to home. L257 which is a cousin of DF95 in the Z18 project. I’ve taken a snapshot of the draft big Y tree for L257 created by Peter OpdenVelde of the Z18 project. this is not all of it..I just couldn’t fit it all on the screen. Again, looking at age estimates, this would represent a common ancestor in about 1800BC. Roughly 4000 years of L257 branching.


With those trees as a reference, here is my example of a Cumberland tree..or a DF95 tree, with missing branches filled in.


Our first SNP in those top 14 would pop in about 140 years or so after Z18 man. So one of his descendants that might show up at his family reunion probably started it off. So maybe around 2100BC. I’ve used S8387 here, but any one could stand in since we don’t know the order of them. I’ve drawn in various branches that might exist but are missing for many of those 14 SNPs. These are the broken branches in our tree. We should see many branches, but our tree is more like the Charlie Brown Christmas tree. All of those branches should be there, but they aren’t. We’re left with our single branch related to a man born around 600BC (based on age estimates). So here age estimates would put the gap around 1500 years, not too far from the 2000 year estimate based on the average 140 year per SNP rate.

Where did they all go? What happened to them? Did we face some sort of extinction event that ended around 600BC? What can we know about Mr. ZP86 (A.K.A. the guy who lived)?





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.