Wandering Thoughts on Ancient DNA

A fellow ZP121 person recently asked how many ZP121 men I thought might exist in the world. It’s a tricky question because there are about 57 men known now. Each might represent a family with tens or hundreds of living men.

Y DNA testing is kind of a rich man’s game and a niche pursuit. Coverage isn’t even across populations. It favors Europe, particularly the British Isles and the U.S. The isles are dominated today by haplogroup R, which gives that group an outsized footprint I think. The massive amount of R testers skews things. In the scope of haplogroup R, we’re tiny, but a lot of groups seem tiny when you compare them to R results. It all depends on where you want to start comparing.

All of that got me thinking about attempting to put our one known R-ZP121 ancient DNA result in context with other ancient DNA results from CTS12023/R-DF95, R-Z18, R-U106, and other Y haplogroups. Since this is the first time we’ve appeared in ancient DNA, I’ll start with the Anglo-Saxon study results and do a loose survey of ancient haplogroups and modern testers.

When I’m listing results, it’s good to keep in mind that there may be overlap. For instance, we may have one man left at R-Z18. He really falls well below that, but his DNA is degraded enough that you can’t go any further down the branches. R-Z18 has 2221 testers today. That number contains 149 CTS12023/DF95 testers, but I’ll be listing them separately. The CTS12023 listing contains R-ZP121 to the tune of 57 testers. This will be the same for all the haplogroups. There is some nesting because not all the results could reach a meaningful terminal point in the Y DNA tree. The numbers and locations of modern testers are meant to give an idea of the size of various groups and some major locations where they are found.

Hatherdene Close

map of ancient graves at Hatherdene close
lovingly borrowed from https://www.nature.com/articles/s41586-022-05247-2

Starting with the community for our ZP121 guy HAD005 there are ten men in the anglo-saxon cemetery study in Hatherdene Close. Not a huge sample. In that group, there is one other R-Z18 man down the larger R-Z17->R-Z372 branch of R-Z18. He is HAD006 reported as R-S4031. Family Tree DNA says that branch formed around 1250 BCE. Today there are 332 men under R-S4031. Most are from Sweden, followed by Norway and then Finland. HAD006 is not listed as Continental Northern European (CNE) like HAD005. HAD006 is listed as Western Britain and Ireland (WBI) with 82% WBI DNA and 17% CNE DNA. HAD006 shows no NOR DNA (Scandinavian) but does register some Continental Western European (CWE) DNA. HAD006 is given an older date than HAD005. HAD006 has no grave goods.

Only one Z18 man in Hatherdene is entirely CNE.

Hatherdene has four U106 men total including HAD005 and HAD006. HAD009 is R-FGC53757 (formed around 150 CE) with 32 downstream modern testers. Most of them are from England. HAD009 is listed as CNE with no NOR or appreciable CWE or WBI, much like HAD005. HAD011 is R-FT83328 (formed around 550 BCE) with 17 modern testers, mainly from Scotland, England, and Northern Ireland. HAD011 is all CNE. These two have grave goods.

Less than half the men are U106.

HAD017 is R-S1205 well down a brother branch to U106 formed around 550 BCE. There are only four modern testers in this branch from England and the Czech Republic. HAD017 is roughly a 60/40 split of CNE and WBI. He is not listed as CNE though because he falls below 70 some-odd percent. He has grave goods.

HAD001 and HAD018 are both down the P312 brother branch to U106 in R-FGC33840 formed around 1500 BCE. Although R-P312 as a whole is much larger than R-U106, this branch only has 3 modern testers from Denmark, Ireland, and Norway. It looks like HAD001 and HAD018 may be identical twins…or the same person. Both are a 30/70 split of CNE and WBI with some amount of CWE DNA as well. HAD001 has grave goods. That wraps up Haplogroup R.

The final three men are under Haplogroup I. HAD003 is I-M253 formed around 2250 BCE. This is a major branch of I with nearly 25000 testers (by comparison U106 has 19000 testers), in large numbers from England, Sweden and Germany. HAD003 has some CWE ancestry. HAD003 is almost completely WBI and has grave goods.

HAD015 and HAD016 are both I-A12775 well down from I-M253 on a branch formed around 50 CE. there are only 12 modern testers from Germany, England, and Russia. Both HAD015 and HAD016 are completely CNE with no CWE ancestry and have grave goods.

Ok. we have ten men. Five are CNE (Continental Northern Europeans). Two of those are Haplogroup I and three are from U106. Of the R-U106 CNE men, the major branch winner is R-L48 with two men to our one R-Z18->ZP121 man. Five men are WBI (Western Britain and Ireland). One of them is from Haplogroup I. Two from R-P312. One is down a brother branch to R-U106 and R-P312, and one is U106->Z18.

In Hatherdene, Haplogroup R is the clear winner. R-U106 has a slim majority of CNE men over I-M253 branches with R-P312 a plurality in the WBI camp if the two people are actually twins and not the same person.

Haplogroup I and R-U106 dominate the CNE group but also occur in the WBI group.

My Bias

I’m biased toward thinking the single R-Z18 WBI member in Hatherdene is descended from an earlier CNE man. When looking at other 50/50 or 60/40 splits with CNE/WBI DNA it’s easy to imagine a WBI father and CNE mother or in the case of HAD006 a CNE father and WBI mother. Just one more generation of WBI ancestry would change the ratio to 20-someodd %. So it seems plausible that HAD006 could be the grandson of a CNE man and a WBI woman and the son of a 40/60% CNE/WBI father and WBI mother.

That seems possible, but I also have to acknowledge that I’m biased to think that R-Z18 favors the Baltic and Scandinavia before the migration period and that there are other possibilities.

Zooming out to Cambridgeshire

Map of the Fens including Cambridge

Including Ely, Hatherdene, Oakington, and Linton.

There are 24 men (including the Hatherdene men above. 11 men are in haplogroup I. 13 are in Haplogroup R. In Haplogroup R, 9 men are R-U106. Three of the men are R-P312, and one is R-S1196. In R-U106, 5 men are R-L48, 2 men are R-Z18 (both listed above), 1 man is R-S1688 (also listed above). There are two WBI men, one in Haplogroup I and one in R-Z18 (mentioned above).

In Haplogroup I, there are 8 CNE men under I-M253, I-BY1330, I-A12775, I-S27836, and I-FGC69701. There are 3 admixed men left at the major branch I-M253, and 1 I-FGC69701.

I was curious about I-FGC69701 which appears in CNE men and in an admixed man, and I found that OAI008, who is not listed as CNE, is nearly completely CNE with a small amount of WBI DNA and some Norse DNA. Kind of light admixture.

In Haplogroup R, there are also 8 CNE men all under R-U106. 5 of those are under R-L48 (R-S19342, R-S21607, R-Z330, R-BY3730, R-FT83328) and then 3 under other U106 branches (R-FT183222, R-FGC53757, R-ZP121).

Admixed men have 3 under R-P312 and 1 under R-Z18 (HAD006 above) and 1 under R-S1205. All of the P312 tests fall under R-L21 which is a major group in R-P312 with roughly 30,000 modern testers.

R-L48 is a major haplogroup under R-U106 with almost 9000 modern testers. Three of these R-L48 men fall under R-Z9 with about 6000 modern testers. Again a pretty large and old group that is thought to have formed around 2150 BCE. Two of the R-L48 men fall under R-S23189 which only has 343 modern testers, but still formed a long time ago in 1950 BCE.

Again, Haplogroup R for the narrow win, and R-U106 is popular among CNE individuals. R-Z18 is trailing in U106 with one CNE and one WBI.

Haplogroup I and R-U106 are on top of these Cambridgeshire burials for CNE with R-P312 men popular in the Admixed group.

My Bias

Because R-P312 is often compared to U106 as Celtic world vs. Germanic world it’s easy to put the Admixed CNE/WBI R-P312 men in this category as WBI men paired with CNE women, which may be the case. It doesn’t have to be the case though. R-P312 has more Big Y testers from Germany than R-U106 does. It’s not that R-U106 dominates the Germanic world; it’s that the Germanic world dominates R-U106. R-P312 is bigger and has more German people, they’re just outnumbered by British and Irish testers in R-P312. In that way, four percent of R-P312 is actually bigger than eleven percent of R-U106. Those R-P312 admixed men could have had CNE male ancestors for all I know.

The Haplogroup I-FGC69701 person listed as OAI008 who is nearly all CNE with other CNE men falling into the same haplogroup, makes me think his paternal line is CNE with light admixture from a WBI grandparent. I really don’t know that much about Haplogroup I other than Family Tree DNA once called it Europe’s native son and that it appears to be a dominant haplogroup in older European burials. I’m not surprised to see it in both CNE and WBI people, and I’m not familiar with whether one branch leans more continental while another leans more towards Britain. So I’m biased here by the other CNE results in the same haplogroup.

Anglo Saxon England

Map of Britain circa 600 AD with labels for anglo saxon kingdoms

Whew. Alright, after some shuffling and sorting I think I’ve got Anglo-Saxon study results without vikings, iron age samples, bronze age samples, samples from the continent etc. This should include more cemeteries with some WBI results.

The Outliers

Haplogroup E: Two samples. One is E-P177 (admixture unknown). The second is E-CTS5856 and He’s CNE and he’s from Eastry.

Haplogroup G: One sample G-P303 (admixture unknown), location unknown.

Haplogroup J: One sample J-CTS5789 from West Heslerton. This one is CNE with some Norway DNA.

The Big Two

Haplogroup I:

There are 52 samples. Four of those samples are WBI. Three samples are 50/50 or 60/40 CNE/WBI. Two of them are 80 or 90% CNE. One is unknown. The other 43 are Continental Northern Europeans. The Majority were I-M253. Of those most fell into I-M253 > DF29 > Z58 and below. That group has roughly 7000 modern testers mainly from England and Sweden. A decent chunk fell under I-P222 > CTS616 > CTS10057 > Z161 > CTS4348 > L801 which has about 3000 modern testers mainly from England and Germany. Smaller amounts of men fell under I-S2488 (60 modern testers from England and Ireland) and I-L38 (830 modern testers from England and Germany).

Haplogroup R:

There were 81 Haplogroup R samples. fifteen samples were WBI. Twenty-one samples were admixed CNE and WBI with some people in the high 80’s and 90’s as a percentage of either CNE or WBI. Two samples were unknown admixtures. The other 43 men were Continental Northern Europeans. Which is kind of ironic that it matches the CNE mix from haplogroup I.


Under R-U106 only one man was WBI (our Z18->Z17->R-S4031 friend mentioned in Hatherdene). The rest of the U106 results were either Admixed CNE (7 men) or CNE (32 men). There were 19 R-L48 men. There were 8 R-Z18 men. Four of those were R-DF95/CTS12023 which is pretty wild because we’re 7% of modern R-Z18 testers, but 50% of the R-Z18 results in Anglo-Saxon England. One man was R-S19726, from Buckland cemetery in Kent. Admixed nearly 50/50. You may recognize R-S19726 from this old post about DF95’s hidden brother in the R-Z18 tree.

For admixed U106 men, we have R-FGC23205 (popular in France, England, and Germany), R-S19726 (mentioned above. popular in England and Sweden), R-Z381 (a major group of U106 with 14000 plus testers, popular in England and Germany), R-BY3326 (popular in Germany and England), and R-BY20443 (popular in England, Germany, and the Netherlands).

For CNE U106 men, we have quite a list. I’m leaving the U.S. out because it’s a given that many modern testers are from the U.S.:

  • R-S18632 (England and Netherlands) 95 testers.
  • R-Z381 (mentioned above the major group under U106 14000 plus testers)
  • R-A6707 (England and Germany) 7 testers
  • R-S19342 (England and Germany) 105 testers
  • R-S21607 (Finland, Scotland, England) 60 testers
  • R-Z9 (England and Germany) 5765 testers
  • R-Z330 (Germany and England) 1309 testers
  • R-BY3730 (Scotland and Germany) 30 testers
  • R-PF740 (England and Germany) 58 testers
  • R-Z345 (England and Scotland) 4002 testers
  • R-Z2 (England and Scotland) 3975 testers
  • R-CTS10893 (England and Scotland) 582 testers
  • R-FT83328 (Scotland, England, Northern Ireland) 17 testers
  • R-FTB13672 (Germany and Ireland) 16 testers
  • R-FT183222 (Netherlands and Ireland) 6 testers
  • R-DF94 (England) 262 testers
  • R-FGC53757 (England) 35 testers
  • R-Z18 (England and Sweden another major branch of U106) 2221 testers
  • R-PH1163 (Denmark and Norway) 2 testers
  • R-CTS12023 (England and Germany) 149 testers
  • R-ZP121 (England and Germany) 57 testers
  • R-BY13800 (Sweden, Finland, Germany) 16 testers
  • R-BY50725 (England, Ireland, Germany) 21 testers
  • R-BY62920 (England) 5 testers
  • R-Z154 (England and Ireland) 85 modern testers.


Two WBI men were R-P312. The majority of WBI men were left back at R-P310 or R-M269, basically undetermined.

Roughly 14 R-P312 men were either CNE (8 men) or admixed CNE (6 men). Among WBI P312 men we have R-DF41 (most popular in Scotland and Ireland) and R-Z253 (most popular in Ireland and England).

For Admixed P312 we have R-FGC33840 (mentioned above), R-FT130235 (most popular in England and Scotland), R-BY35104 (most popular in England), R-CTS11567 (most popular in England and France), and R-BY61198 (found in England).

For P312 CNE men we have R-CTS11567 (most popular in England and France), R-Y31393 (most popular in Portugal, England and Mexico), R-Z274 (most popular in England and Spain), and R-BY31939 most popular in England.


There were about 17 R-M269 men left above the P312/U106 groups. They mostly belonged to WBI and admixed men, but some were also CNE. These samples are probably degraded and can’t be driven any further down the tree.


There were 3 R-M198 men (R1a to our R1b), one each of Admixed CNE/WBI, WBI and CNE. R-M198 appears to be most popular in Poland, Russia, and Germany.

My Biases

Big vs. Small

u106 visual tree
borrowed from http://www.jb.man.ac.uk/~mcdonald/genetics.html

I’m in haplogroup R with 105,530 testers. By comparison, Haplogroup I has nearly 41000 testers. More specifically, I’m part of R1b (R-M343), which may still be the most common Y Haplogroup in Europe. There are almost 88000 testers in R1b. Haplogroup I-M253 has about 25000 testers. My perspectives are skewed by being part of these massive Y groups. What I consider large or small groups are based on comparisons to other R haplogroups. My small groups may be absolutely massive.

That skewed perspective comes into play when I say I’m in a smaller R1b group; R-U106 (18,810), because I’m comparing it to R-P312 (52,789). Further, when I compare the size of my group R-Z18 (2221) in my head, I’m comparing it to R-Z381 (14,603). R-Z18 is “smaller,” but there are other groups on the same level that are much smaller than that. R-A10122, a sibling of R-Z18 only shows 7 testers at FTDNA.

When I say my group under R-Z18, R-CTS12023 (149) is small, it’s because I’m comparing it in my head to R-Z17 (1,187). The truth is there are several smaller and rarer groups under R-Z18 now. We’re “smaller” than the bigger branches of course, but not the smallest.

That is why it doesn’t surprise me that there are 27 R-Z381 results in the Anglo-Saxon study. That group is larger today so it seems likely there would be more of them. If R-Z18 has roughly 15% of the number of modern men that R-Z381 has, then I’d expect there to be about 4 R-Z18 results. Since there are 8 R-Z18 men, double my expected amount, that is interesting. Then thinking about CTS12023 sitting at about 7% of modern R-Z18, but 50% of the Anglo-Saxon study R-Z18 men is pretty amazing to me by comparison.

YDNA and Autosomal DNA

One of the reasons Y DNA works so well for tracing long lines of ancestors is because it is relatively stable. It might be passed on in exactly the same form for multiple generations before a mutation pops up. Autosomal DNA changes pretty dramatically every generation, and half the input is female. Moms matter. A Y DNA origin doesn’t have to line up with autosomal DNA…really in any way after a few generations. When I see CNE men turning up R-Z18 that fits my bias, but when I saw a WBI man turning up R-Z18 it conflicted with my bias, and I constructed a way to explain it that didn’t include an Isles origin. That person’s life may have been completely defined by being British, no matter what I think about his Y DNA and a small percentage of CNE ancestry. When I’m interpreting results, I’m also playing to my own expectations.

ancestry.com map showing U.S. groups in my autosomal DNA

As an example; Although I can trace my Y DNA to England and then the European continent and at some point the steppe and back to Africa, these DNA communities defined by Ancestry.com are more informative about my Autosomal DNA and who is most likely to match me as a whole person, not just on one chromosome. It’s likely that I share cultural bonds, life experiences, and language along with DNA with these groups in the last few hundred years.

The Only Ancient R-CTS12023/R-DF95 DNA So Far

Right now, this is it for ancient CTS12023, and the results are in some ways, misleading. What I get from the study is that people from Belgium up through the Netherlands, Germany, Denmark, Norway, Sweden, and touching on the Baltic seem to define the Anglo-Saxons buried in England and R-CTS12023 in that era. I think that matches our modern test base in a lot of ways too although we’ve obviously moved a lot farther afield.

It is important to note that this map of Continental Northern European sources currently fits all of the CNE people, in every haplogroup, with every Y DNA and mtDNA branch attributed to CNE people in the study. This map is based on autosomal DNA. Our closer Y DNA cousins are a small part of that larger group.

continental sources for Anglo Saxon DNA showing sites in the low countries, northern germany, denmark and sweden.
lovingly borrowed from https://www.nature.com/articles/s41586-022-05247-2

Our CTS12023 cousins have an outsized contribution to the Anglo-Saxons in this particular study that you don’t see in the ratios of modern testers. We’re not a “large” group in the study, but larger maybe than we ordinarily would be and that is mainly because of samples in Anglo-Saxon Kent.

Getting back to the original question about ZP121, so far, there is only one lone Angle from Cambridgeshire out of something like 145 men, in what you could now consider our target demographic.

Looking at the results, I think we see R-ZP121 because this study was so specifically targeted to this group at this time, and better more granular testing was available.

I think it’s time to zoom out again and look at that wider ancient world that we haven’t yet appeared in.

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.