Hartley DNA & Genealogy

I was happy to see recently that Robert Casey has done a study on Z16343. He was primarily looking at YDNA STRs which are better tested but less predictable than SNPs. Z16343 is important for my branch of Hartleys as it is under that branch.

Where is Z16343?

Z16343 is under R-L513, which is under L21 which is under R1b. Here is page 1 of 2 of R-L513:

If the two pages were to be put together, Z16343 would be in about the middle of L513. The circle indicates where Z16343 is. It is a fairly small branch. The names circled represent only the branches tested by the BigY SNP test. Here is a closer look at Z16343:

Here I see only the following 13 surnames:

1. Pilsbury
2. Smith
3. Hays
4. Goff
5. Merrick
6. Hartley
7. Hays
8. Davis
9. Williams
10. Martin
11. Thomas
12. Philips Bennett
13. McCullers

These surnames are in 11 Haplogroups.  Four Haplogroups contain two surnames:

1. BY11565 – Goff and Merrick
2. FTB44077 – Davis and Willliams
3. FGC33966 – Martin and Thomas
4. FT135561 – Bennett and McCullers

There are two likely reasons that there would be more than one surname per haplogroup. One is that the haplogroup formed before the time of surnames. When surnames were adopted, different surnames would have been used for people in the same haplogroup.  The other reason is due to a NPE, a non-paternal event. This would be due to an adoption or single mother situation for example where the male surname would have changed over the years.

Due to additional testing, two Surnames have additional branching:

1. Hayes – ZS349 and BY13845
2. Hartley – A11132 and FT225247

Under Hartley, where I have particular interest, there were four testers. I tested my brother to get a terminal haplogroup named. There are two other Hartleys that are under A11132.

There are three empty haplogroups. These are at branching points:

1. Z16343
2. Z16854 (the side with Pilsbury and Hays)
3. Z17911 (the side with all but Pilsbury and Hays)
4. FT9480 (a branch with all but Pilsbury, Hays, Smith and Hartley)

SAPP

Robert Casey uses the SAPP program for his analysis. Here is the SAPP web page:

The program is by David Vance. Here is Casey’s results for Z16343:

This tree is larger than the BigY SNP tree, because it includes more testers. Casey’s report says:

Z16343 has 19 testers that were confirmed to be Z16343 via YSNP testing. Another 59 testers were predicted for a total of 75 testers. 

I have not used the SAPP program myself. I notice that it has a SNP Tree function, so I tried that for Z16343 and came what looks to me to be an upside down tree:

I’m used to having the older branches at the top, but this one has the older branch at the bottom. When I add names, I get this:

When I compare this tree to the L513 Tree, I see that Davis and Williams are missing in a parallel branch to Martin and Thomas in the above tree. Also Hays at ZS352 is ZS349 in the L513 Tree. The thing I don’t like about this tree (and other trees) is that it seems to imply that Hartley descends from Smith. In actuality, Hartley and Smith descend from a common ancestor who lived before the time of surnames. According to YFull, A11138 formed 1750 years before present or around the year 272! However, YFull does not have the full results of those who have completed the BigY test.

More on Casey’s SAPP Tree

Casey mentions the signature for Z16343 and I think I can pull it off the tree:

I will ignore the ReclOH as I don’t understand that very well (sorry McCullers). One problem with the SAPP tree is that is seems to imply that the McCullers Branch is directly under Z16343. However, by the SNP tree it is really Z16343 > Z17911 > FT94840 and so on. Jared Smith has recently drawn his own tree which includes the parent SNP to Z16343 which I have copied below:

The first branch after the McCullers Branch has Pilsbury (shown above) and Hays (not shown). I’ll put the STRs from above Node #99 (or BY13850} into a small spreadsheet:

Not counting the ReclOH, there are 7 main branches under Z16343 on the SAPP generated YDNA Tree. Node #150 is one of those 7 major branches, but it has no STR deviations from the signature for Z16343. However, it’s two sub-branches have deviations

I’ll show them like this in my spreadsheet:

So far I have 9 STRs in the Z16343 Signature. Casey said that there were only 8 STRs in the signature:

The signature of Z16343 includes eight markers which is acceptable. But this signature does not include slow mutating markers or multi-step mutations and is not the strongest signature.

Here is my spreadsheet for the seven main branches under Z16343:

This shows 12 STRs in the signature. I added in dates for the branches. Pilsbury has the latest date for BY13850, but also the most variations from the signature Z16343 STR. I highlighted the two 12’s under 439. That could normally be a problem, but these were apparently determined to be parallel mutations as the SNPs showed that they were in a different branch.

The Hartley Branch of the SAPP Tree

As I know most about the Hartleys, I will look at what Casey’s SAPP Tree did for them:

I have started by highlighting my branch which I feel cannot be right. My brother and I have the obviious ancestor of my father. He was born in 1918. The date given is 1900 which is not far off. The 1-3 generations is right. However, it is the relationship with Steve that seems off.  Here is my brother Jim’s block tree:

This shows that Jim and I have 7 SNPs up to our common ancestor. Steve has 5 Private SNPs but only 2 average SNPs. If I say there are about 5 SNPs up to our common ancestor and I use 83 years per SNP, then that is 415 years. If I take that time from 1960, then that goes back to the year 1545. Another check is Michael’s genealogy. His ancestor Edmund Hartley was born in 1666 in Lancashire County, England and came to Pennsylvania around the year 1700. His father Roger Hartley was born in 1628. I don’t think that I descend from Roger, but if I did, that would mean if the block tree is right, I couldn’t have a common ancestor with Steve and Michael any later than 1628.

The SAPP Tree and Steve

The SAPP Tree shows a much later date for the common ancestor between me, my brother and Steve. By this I mean much later than seems to be warranted given the YDNA SNPs:

That shows a range between 1750 and 1850. That could be off by 3-400 years. Both the SNPs and our genealogy suggest that 1800 cannot be right.

The SAPP Tree and Michael

However, the SAPP Tree for Michael shows a common ancestor which appearst to be too far away.:

Node #121 shows a date between 1200 and 1550. The SNPs seem to suggest that both Michael and Steve and I should have a common ancestor around Node #101 or around the year 1500.

The Other Three Hartleys on the SAPP Tree

I’m curious as to how the SAPP Tree would have come out with John’s new 111 STR results. Here, he is shown as further from me compared to Steve. John is my closest match other than my brother at the 111 STR level:

The TiP report gives a date to a common ancestor with John at about 1530, so that part is not so far off. It is more that Steve is shown as matching more closely than he should be. The SAPP tree also gives a common ancestor between me and John at the year 1500 which seems reasonable:

Lawrence has recently put in for a BigY 700 test, so he is a very important YDNA tester.

My Previous Attempts at Hartley STR Trees

I have made a few attempts at drawing my own Hartley STR trees. Here is one that I built:

This looks somewhat similar to the SAPP tree. This reminds me of my frustration at having Steve test and finding out that he was not related as closely to me by SNPs compared to what the STRs seemed to show. The common answer to the discrepancy lies in back mutations and parallel mutations which are difficult to detect. In my tree, Michael is on the right and in the SAPP tree, Michael is on the left. However, the tree is basically similar. In the tree above, I tested as well as my brother. Interestingly, I had a new STR mutation in 1956 when I was born at STR 534. Steve has completed his BigY test. Lawrence ordered a BigY recently as did John. However, I don’t see John on my chart. That may be because he tested to 67 STRs and this could be a 111 STR chart.

John is on the SAPP Tree:

The tree shows our common ancestor at Node #101. However, once Steve’s BigY information comes in, it will clear up the tree.

This is the SNP working model I am using:

In this model, with the information I now have, we are about at the same level for a common ancestor. My guess is that our common ancestor is around 1500. Actually, that is true for all but John. John and Lawrence are new testers, but Lawrence’s dating should be tied to Michael’s.

More on John’s STRs

John’s STRs were updated recently as part of his BigY order. I wrote a short Blog on those results here. My best guess using FTNA’s TiP Report is that John and I should have a common ancestor around the date of 1565 (although I also have a 1530 estimate above). The interesting this is, that by STRs John is related most closely to my brother Jim and then me. This is because I had an extra STR mutation that my brother Jim did not have. Then after that, John shows a more distant match with Steve. There are more Hartleys out there that have tested and are related to us by YDNA, but the STR test does not show a match with them because they are beyond the FTDNA matching limit. That includes Michael who has taken the BigY test.

Summary and Conclusions

• In my Blog above, I took a rambling look at Casey’s SAPP tree analysis of Z16343
• I don’t believe that the SAPP Report that Casey did was meant to supersede work already done with trees based on SNPs
• Casey’s main point was that STRs should be used to increase the pool of predicted families under Z16343. Based on his study, he went from 19 testers under BigY testing who would be under Z16343 to 75 who should be under Z16343 by STRs.
• The tree works out overall, but in the details where there are discrepancies between the SAPP STR tree and the SNP tree, the SNP tree should be used.
• Every study and analysis that is done on the exisitng YDNA testing brings a little more understanding and light to the architecture of the Z16343 tree.

I was glad when John decided to take a new BigY test. I see that the first part of John’s BigY test is in. That is his 111 STR results. Previously, John was tied as the closest with Steve to my brother Jim at 67 STRs:

Jim matches Steve and John at a GD of 3. Michael doesn’t show perhaps due to his match settings or the Genetic Distance was too far.

At 111 STRs, Jim matches John a little more closely than Steve by STRs:

Here John is seen as matching more closely by STRs.  However, in the past, Steve seemed to match me better by STRs, but when Steve took the BigY SNP test, it appeared that Michael, Steve, my brother and I all matched at about the same level.

The TiP Report

My brother Jim has this TiP Report with John:

I’ll use 90% confidence and I get 10 generations. I’ll use 35 years per generation and get 350 years. I’ll use 1950 for the present generation which comes out neatly at the year 1600. That comes out closer to the present than the year of 1500 I was using for the common ancestor of Michael, Steve and myself.

The TiP Report with Steve

By comparison, here is Jim’s TiP Report with Steve:

This brings us back another 2 generations or another 70 years to 1530. This seems consistant with the SNP analysis that I did previously. The reason that I am using Jim’s results is that I have an extra STR mutation that Jim does not have. Now the fact that I have this mutation could mean that one was due, or that because two brothers tested, there was more chance of getting that mutation.

I suppose that it would be more accurate to do the calculations from my viewpoint also and to average the two dates.

TiP for Joel

With Steve and myself I get an estimated common ancestor at 14 generations with a 91% probability. That averages out to one more generation or to 1495 or even closer to the year of 1500 I was using with the SNPs.  Between John and me, it is 12 generations at 91% . If I were to average the estimated date of the common ancestor between John and myself and with John and Jim, I would subtract another generation to get the year 1565.

STRs from John’s Point of View

From John’s viewpoint, the difference is greater:

The difference between John and my family is that I have that extra mutation that Jim did not have. I already looked at the TiP report between John and Jim and John and myself. Here is the TiP report between John and Steve:

Here I will use 1950 – (16 X 35) or 1390. That is a long time ago to have a common ancestor if that is right. Those must have been some early Hartleys. Based on STRs alone, it would appear that My family is slightly more closely matched to John than to Steve and that Steve is more closely matched to our family than to Steve’s Branch of Hartleys.

John’s 67 STR Results

The 111 STR test is more accurate. Mervin fell off the list at 111 STRs as the matching was more distant. Lawrence has ordered the BigY test, so it will be interesting to see how the different Hartley YDNA tests compare with each other.

I will be looking forward to see how the rest of John’s BigY test turns out. I could try to draw a STR tree, but my past attempts have been a lesson in futility, so I will forego that exercise at this time.

Here are current lab times:

Summary and Conclusions

• The first step of John’s BigY test is completed. However, the significance of the STRs should pale compared to the results of John’s BigY SNP testing
• Based on the TiP report John should be related more closely to my line than other Hartleys who have done the 111 STR test
• Based on the same report John has only three 111 STR matches. Those are all Hartleys, but two of those matches are my brother and me. The other is with Steve. The TiP Report implies that John’s match with Steve is more distant than with my line of Hartleys.
• Time to wait another month or so for the rest of John’s results

The good news is that two more people in my Hartley A1132 Branch have ordered the BigY 700 test. This is the current state of BigY testing for my Branch of Hartleys:

There have been four BigY testers. This simple tree is a mixed tree as my branch (Joel and Jim) should go back to Robert Hartley born about 1803 probably in Trawden, Lancashire, England. Here is an updated tree to include the new testers:

Here I ordered the different Hartley Branches by ages of oldest Hartley ancestor. We are still working on pushing these dates back on these genealogies. There are candidates for older ancestors, but it is difficult to be sure the further back we go. The new testers are Lawrence and John. It will be a while before their results are in. My branch is the only one with a new SNP. That is because I had my brother tested using the BigY 700 test.

Dating FT225247

There are many ways to date SNPs. One simple way is to use SNP tracker. SNP Tracker is a website and has Hartley ancestors going back to genetic Adam:

I cut off the scale on the bottom. The purple color goes back to the Paleolithic Age. Here is a more close-up view:

I expect that the FT225247 red dot would be more accurately further North in Lancashire. Here is some more information from SNP Tracker:

The little skull indicates that there is some verification from an archeolgical site. The blue cross to the right of R-A11138 is for Scotland. This match is with a Smith and goes back before the time of surnames. A11132 is dated at 850 CE. This should also be before surnames. That means that, unless all the other surnames died out, there could be other surnames associated with A11132. So far only Hartleys have been tested and have been found to have this SNP. FT225247 is intersting because SNP tracker gives a date of 1500 CE for this SNP. I don’t totally understand this date because the number before that is 440 years before present. As it is now close to 2022, that should be 1582. I don’t understand the difference of 82 years. One thought is that if both my brother and I have this SNP, then my father must have had it also and he was born in 1918.

The YDNA Block Tree

My Block Tree shows those who have taken the BigY test:

The specific area that I am interested in is where the arrow is. Above that space, all A11132 Hartleys have those 10 SNPs in common. Below that space, Jim and I have 7 SNPs in common. The 10 SNP block is represented by A11132 and the 7 SNP block is represented by FT225247. That means that the specific SNP FT225247 (my Hartley Line) may be as old as 1500 (assuming SNP Tracker is right) or as recent as 1918 when my father was born.

Likewise, the specific SNP A11132 may be as old as 850 CE (using the SNP Tracker number) or as recent as one generation before FT225247 or say, 1465.

The Iain McDonald SNP/STR Method

A specialist in the field of YDNA came up with this chart for TMRCA or time to most recent common ancestor:

The SNPs are in the vertical column and the STRs are in the horizontal row.

My Brother Jim and Me

Between my brother and myself, for example, there are no SNP differences, but there is one STR difference. That puts the time to most recent ancestor (our father) at 5-141 years ago. One of the instructions says:

- To convert these to date of your MRCA's birth, subtract these numbers from your averaged birth dates. If this is not
  known, a good approximation is 1956 AD/CE. Hence, a genetic distance of 2/111 with 3 mis-matching SNPS should be read as
  48-313 years. This becomes 1643-1908 AD at 95% confidence.

I was born in 1956 and my brother was born 1949, so say 1953. That means the ranges are between 1812 and 1948. As I mentioned above, my father was born in 1918. I was a bit surprised to have a STR difference with my brother.

Steve and Me

I think that Steve is a bit younger than me, so I’ll say our average birth age is 1960. Another instruction for the chart says:

- SNPs should be counted consistently with coverage of the test. There is no one-to-one correspondence possible here for
  any company without going heavily into the raw data. For Family Tree DNA customers, the number of SNPs that is
  appropriate will normally be somewhere between the number of non-matching variants and the sum of both sets of private
  variants.

You may have to scroll to see all the instructions. It says the number is between the non-mathcing variants and the sum of both sets of private variants.

This shows that I have 15 non-matching variants with Steve. Interestingly, I have one with my brother Jim.

By sum of Private Variants, I’ll add my SNPs (7) to Steve’s Private Variants and get 12. That means that between the two is 13 or 14. As far as STRs go, Steve and I are 7 steps apart. If I am interpreting the instructions correctly, I get this:

I suppose that I could average these numbers and come up with 308-837. That is quite a spread. That means that our common ancestor lived between 1123 and 1652.

Steve and Jim

I’ll check to see if there are any differences. Jim also has 15 non-matching variants compared to Steve. However, Jim and Steve differ by 6 steps on their STRs. That makes them a little more closely related by DNA (though in reality, there should be no difference). In the chart above, that shifts the numbers overto the left by one column. So 280-812 years from 1960 or 1148-1680.

Counting Block Tree SNPs

Tiger Walsh says on the same Facebook site where he introduces the above chart:

The problem with this is that my Block Tree shows 7 mutations in my line and only 2 average mutations between Steve and Michael. I believe that part of that problem is that Michael had the BigY 500 test and missed some of his variants. My best guess is that there should be about 5 variants between us. If I take 5 variants at 83 years each, that comes out to 415 years. That compares well to the SNP Tracker number of 440 years ago.

If I apply the 83 years to the top part of the block tree, that is an additional 830 years.

This is what I end up with. However, I don’t know what to use for a starting date. I’ll use 1960. That gets the common ancestor between me and Steve at 1545. That puts the top of the chart another 830 years back or at the year 715. Compare this to the SNP Tracker which had:

Mushing all that information together I get:

That would put the horizontal space inbetween the upper block and two lower blocks at about 1500. Based on the incomplete testing to date, that is probably the best date that I can come tofor the common ancestor of the existing three groups of Hartleys. Fortunately, all this information will be updated once we get the results of the two new BigY A11132 Hartley tests. The top block predates the use of the Hartley name and represents the time when the Hartley surname developed.

Dating the Roger Hartley Line

Once Lawrence’s BigY results are in we will have some genealogical correlation with the BigY:

The branch that Lawrence forms with Michael should date to their common ancestor Edward Hartley born in 1666.

Summary and Conclusions

• The existing BigY Tree of A11132 Hartleys had 3 branches. The new tree will have four testers plus an upgraded branch for Roger Hartley whose genealogy of all the Hartley branches goes back the furtherst (to 1628)
• I reviewed the SNP Tracker dating of the A11132 and FT225247 Hartley Branches
• I looked at a new Iain McDonald method of dating TMRCAs by using SNPs and STRs
• I looked at using existing SNPs and Variants at the rate of 83 years per SNP for comparing the existing Harltey BigY 700 tests. This method compared well with the SNP tracker dates
• I came up with dates on the Block tree for the different Hartley SNP blocks
• I noted the Hartley common ancestor date of 1666 for new BigY 700 tester Lawrence and existing BigY 500 tester Michael
• The A111132 Hartley YDNA Tree should be in much better shape when the two new Hartley BigY tests are completed

This Blog is a follow-up on my previous Blog. That Blog was about a new A11132 Hartley BigY tester. It was also about his genealogy and the genealogy of the other three existing A11132 Hartley BigY tester groups. I had this simple graphic for the existing situation of A11132 Hartleys:

The new tester was not included in this graphic as his test will take six to ten weeks to finalize according to FTDNA.

Finding Answers As To Why the Roger Hartley Branch Has No Private Variants

That is one of the questions that I am  trying to answer. Here is my Block Tree:

I have 7 SNPs. Steve has 5 Private Variants and Michael has no Private Variants. I would like to know why in around 500 years, Michael shows no Private Variants.

Test Coverage

My understanding is that Michael took the BigY 500 test. That seems to be true based on this comparison:

Michael has about half as many Named Variants compared to Steve, Joel and Jim. That seems to confirm that he had the earlier BigY 500 test. There are 10 Named Variants per page.

Comparing Private Variants

First, one of my assumptions seems to be wrong. Michael shows two Private Variants:

This does not make sense as Steve has 5 private Variants. The average of 7 Private Variants should be three of four, not two. I asked about this in the FTDNA – BigY Facebook page and got this answer:

My takeaway is that all the Private Variants are used in the individual lists. However, when they average the Private Variants, they only use the best variants to be consistant with the variants used for calling SNPs.

I have started a Private Variant comparison Chart:

In the chart above, Steve has five Private Variants and Michael has two. I checked Steve’s first PV against Michael and Michael’s results were inconclusive. That means that Micjael and Steve may share a SNP in that location.

My brother and I have no Private Variants, but we have SNPs during the time Steve and Michael were having Private Variants:

Hopefully, these SNPs have been checked out, but I will put them in to see for myself.

Comparing Steve’s 5 PVs with Michael;s Results

For Steve’s first two PVs, Michael had a ? in his results. I am not looking at the detailed results, just a summary table. However, that tells me that there is a possibility that Steve and Michael could share results which would mean that they would share a SNP and a new Hartley Branch.

Comparing Steve and Michael’s PVs with My Results

These results were unambiguous:

I clearly had none of the PVs that Steve and Michael Show.

Comparing Michael’s Two PVs with Steve’s Results

In this case also, the results leave no doubt:

It is also possible that one position may be covered in on test and not in the other. However, that has not happened so far, the A11132 Hartleys.

Comparing My Hartley Branch SNPs with Steve and Michael

I’m curious.

The results with Steve are clear:

My Hartleys and Michael

Here are the results for my first Hartley Branch SNP:

Here is my Chart:

I didn’t check my brother’s results as they would likely be the same as mine. I gave Michael an NT twice in the last column for ‘not tested’. I assume that ‘no reads found’ means that FTDNA didn’t cover that position.

By each tester:

• Steve’s Line has had five Private Variants found since the common ancestor of these three lines. However, two are possibly not private as when Michael’s BigY tested two of those positions, the results were inconclusive
• Michael – His test only found two private variants since probably around the year 1500. He should have more, so his test probably missed some. The two PVs that he did have were valid as they were negative in the other two testers
• Joel (and Jim) – They show 7 SNPs since the A11132 common ancestor of the three groups of Hartleys. However, two of those SNPs were not covered by Michael’s BigY test

My initial conclusions are the Michael is the wild card as far as the YDNA testing goes. There are two variants where Michael could forn a branch with Steve and two where he could form a branch with my line. That doesn’t mean that he would form a branch with either, but he couldn’t form a branch with both. That is because Steve and my branch would have to match each other and we don’t. If Michael matched me on the two SNPs marked as NT, that would reduce my Hartleys’ PV from 7 to 5 and match what Steve has. That is to say that if Michael matches either Steve or my Hartley Branch, then it would seem more likely that he would match my branch.

It appears that more testing is needed on the Roger Hartley Line:

I checked with two people on the Roger Hartley Line and one got back to me and said he would look into the test, so that is good news. If the new tester adds his results to John’s then we should get an interesting Hartley tree that will extend back to before what we are to figure out from the Hartley genealogy. The YDNA should fill in some of the blanks that we have in the Hartley genealogy.

Summary and Conclusions

• Michael’s genealogy is most important for an understanding of how the different A11132 Hartley branches fit together
• However, Michael took the BigY 500 test unlike the other A11132 Hartleys who took the newer BigY 700 test. Because of this Michael has some ambiguity in four of his important variants. These variants have the potential to have an effect on the two other Harltey Lines (Joel’s and Steve’s).
• Michael had ambiguous results for two out of the five Private Variants that he shares with Steve
• Joel and Jim are brothers who both tested with the BigY. Their testing resulted in the naming of alll their 7 shared variants.  Michael clearly did not have 5 of those shared SNPs. However, for another two, Michael did not have any reads. That means that we don’t know if he shares those SNPs with Joel and Jim’s new Hartley Branch or not.
• If the Roger Hartley descendant that I have contacted takes the BigY 700 test, he should clarify Michael’s 4 ambiguous Variants. These 4 ambiguous variants have an effect on my line and Steve’s line
• I am looking forward to new A11132 Hartley BigY 700 results.