My Mother’s Rathfelder DNA: An Initial Look

My mother is a Rathfelder. That is a fairly uncommon name. According to forbears.io, it is the 413,549th most common name in the world. By comparison, my last name, Hartley, is in the 6,000’s. My mother’s father Alexander grew up in Latvia. He worked on a ship and jumped ship in New York City in the early 1900’s. So he doesn’t have a lot of relatives around here. My mom’s family tree at Ancestry.com looks like this:

Gladys Ancestry Tree

On the bottom left, are the Nicholson and Lentz families. Ann was born in Sheffield, England. The Lentz family was in current day Philadelphia by the time of the American Revolution.

Here are the DNA testers:

  • Me – tested at all three DNA companies for autosomal DNA. I also tested for mitochondrial DNA. This covers the line on the bottom of the chart only.
  • My 2 sisters – Tested at AncestryDNA and transferred to FTDNA
  • My mother – tested separately at FTDNA and AncestryDNA
  • Judy – a second cousin. She tested at 23andme, but hasn’t uploaded her results yet to gedmatch.com for comparison. Our common ancestors are Jacob Lentz and Annie Nicholson.
  • Catherine – a Rathfelder second cousin in England. Our common ancestors are Joahnn Rathfelder and Maria Gagnus.

I have also test results from my father’s side. When I put all my known test matches together, it looks like this:

Chromosome Map 2nd

The matches we are looking at here are on my Maternal side, so that would be the red (Lentz/Nicholson) and flesh colored (Rathfelder/Gagnus) segments at the bottom of each Chromosome. Note that we receive DNA from both our parents on each chromosome. This means that if someone matches me and Catherine on the same segment of Chromosome 13 that is colored in, I will know that person matches us on my mother’s side. And more specifically, the match will be along one of the ancestors of this Rathfelder/Gagnus couple.

My Mother’s Rathfelder/Gagnus DNA

My mother will have more Rathfelder/Gagnus DNA than I do, because my DNA is watered down with my father’s DNA. If I look at cousin Catherine’s DNA matches with my mother, me and my sisters, this should show us the DNA the 3 of us got from this Rathfelder/Gagnus couple that were born in the mid 1800’s. That’s what I did, and it looks like this:

Catherine Chromosome Browser

Here, my mother, Gladys, is in orange, I’m in blue. My sister Heidi is green and sister Sharon is pink. On Chromosomes 2, 8 and 14, some of the Rathfelder segments didn’t make it to me or my 2 sisters. Also note, that I am missing Rathfelder segments from Chromosome 4, 6 and 10, that the rest of my family have.

Does Anyone Else Match the Rathfelders?

As one might guess, this line is German. I checked at gedmatch.com using a utility showing people that matched both Catherine and my mom, Gladys. Then I compared them in something called an autosomal matrix to see how they all matched each other.

Gladys Catherine Autosomal Matrix

The upper left part of this matrix represents matches between Catherine and my family. The upper right part shows our matches to others that match Catherine and Gladys. The lower right part shows how the people that match Catherine and Gladys match each other. This is important for triangulation and finding common ancestors. From this, we can see that Michael and Tara match each other closely. In fact, they have the same last names, although I don’t show it here. Christine and Kenneth match each other at 20.8 cM, so let’s look at that. It turns out that they don’t match Catherine or my family where they match each other, so there is no triangulation there.

Let’s try something else. On the Excel spreadsheet I have created of my mother’s matches, I show where she has triangulated groups already. Here is the most promising Triangulation Group as seen from Gladys’ matches:

TG Gladys

I went in to gedmatch.com and made sure that Catherine matched the last 2 matches and that the last two matches matched each other. All these overlap on the same Chromosome 13 and at the same areas of that Chromosome. This makes a TG or Triangulation Group. This means that these last 2 people will have the same ancestors as Catherine and Gladys, somewhere up on our shared Rathfelder or Gagnus family tree. Perhaps these 2 people have good family trees and it will be easy to see. Or they may not know much about their family history. At this point, one could contact these people by email to try to find out if they know where we match.

What About the Lentz/Nicholson DNA?

This is a bit more difficult. Cousin Judy has tested at 23andme. I have tested there, but no one else in my family has. I have downloaded Judy’s results to my spreadsheet. There I can check other matches around her matches to see if they match on my mother’s side. Not many do. Here is one spot where there are a few matches at Chromosome #17:

TG Lentz Nicholson 17

The blue matches are on my paternal side. We can ignore those. The pink is my maternal side. The white don’t match either side so are IBC (Identical By Chance) or they go below the threshold when checking which side they are on. At any rate, I can ignore the unhighlighted names for now. LinnyLou and Douglas are closely related. However, they do triangulate with my mom and Judith. That means we likely have a set of common ancestors out there.

DNA the Ancestry Way: Trouble With Schwechheimers

Ancestry automates all this work: easy for me and easy for them. Right? Actually, easy, but not accurate. Here is another mistake I notice that they’ve made. They see where I have a DNA match and then they find if there are common ancestors in our trees and say this is a likely match. The problem is is they don’t triangulate. And Ancestry doesn’t know if our trees are correct. For my Ancestry kit, this is what they found:

Schwechheimer False Match

Ancestry found that I (represented by the line on the left) matched the person represented by the line on the right. We matched at 5.4 cM on one segment. That is tiny, but Ancestry puts the results through a filter, they reason, which filters out the bad matches. So what is wrong? The problem is that I tested my mom and she doesn’t come up with the same match. That means if she doesn’t match, I can’t match with this Schwechheimer as I would have gotten all my Schwechheimer DNA from my mom.

Conclusions:

  • I’m glad to have the testing results of Judy and Catherine because they each represent one side of my mother’s family – Paternal for Catherine and Maternal for Judy
  • Even with these testers, it is difficult to find many matches that triangulate
  • Don’t always trust Ancestry. Upload your results to gedmatch.com where you can see where the match is on the Chromosome and check for triangulation there.

 

3 Generations of Upshall DNA

My wife Marie is not an Upshall. Nor is her mother Joan. However, her mother’s Aunt Esther is. The Upshalls were from Newfoundland and lived in a fishing village called Harbour Buffet. This is what it looked like in 1907

harbor buffet 1907

Harbour Buffet is in Placentia Bay. Here is a map showing Harbour Buffet on the left side in relation to St. John’s on the right. As you can see, Harbour Buffet is quite an isolated area.

Harbour Buffet Map

The genealogical records are sparse in Newfoundland. In Harbour Buffet, the church burned, so vital records are missing.

Are Your Parents Related?

I have had my wife’s, her mother’s and her mother’s Aunt’s DNA tested. Living in isolated Newfoundland many people have ancestors of the same name more than once. I have uploaded these 3 people’s DNA results to Gedmatch.com. They have a utility there called “Are Your Parents Related?”. I ran this and, sure, enough, as Aunt Esther expected, her parents were related. She shows a relatedness on Chromosome numbers 2, 11, 15 and 20.

Related Parents Esther

My understanding is that this report looks at Esther’s mother’s and father’s side of her DNA and where there are matches between the two, it shows where they had a common ancestor. In fact, this report indicates that Esther had a common ancestor 4 generations ago. At 4 generations ago, we had 16 2nd great grandparents. This means that Esther likely only had 14. This could also mean that Esther is a s 3rd cousin to herself! A scenario could be that Esther’s mother and father could be 2nd cousins to each other. I tried to sketch out some of the relationships here:

Upshall Ellis Chart

I am still working on Esther’s ancestry, so I haven’t gotten too much past the diagram above. My understanding is that Melinda Jane Kirby’s mother was a Dicks. That means that her two grandmothers, Catherine Dicks and Melinda Kirby could have been 1st cousins. This is helpful in tracking the ancestry. Dicks is a name that comes up quite a bit when looking at Esther’s DNA matches also.

What DNA Came From Whom

In the chart above, you can see that Frederick Nelson Upshall had 2 wives. The first wife died in the Flu Epidemic in the Boston area and he remarried. That means my wife and my mother in law are descended from the Daley (non-Newfoundland) side and Esther is descended from the [Newfoundland] Shave side. Gedmatch has another report called “People who match one
or both of 2 kits”.  When I run this for Joan and Esther it shows people that match both of them, or just one or the other. The people that match both should correspond to the purple in the diagram above and below. The people that match only Joan and not Esther would represent the left side of the chart. The people that match Esther and not Joan would represent the upper right had side of the chart. Except that there is one hitch. Thanks to some collaboration with at least one helpful DNA match’s research on Ancestry, I have  been able to expand Aunt Esther’s ancestors further back.

X Chromosome Chart Esther

Now we can see that Esther has Dicks family on her mother’s and father’s side. The dotted lines at the top are inferred from the DNA and a guess on my part. I am assuming that the Dicks ancestors are the reason for Esther’s parent’s relatedness. This means that Joan would also be related to Jane Ann Dicks but not as closely as Esther.  Likewise, Joan would be related to Melinda Kirby but not her father John and is related to Margaret Shave (her step-grandmother) but not George Shave. Further, I may assume that the parents of Jane Ann Dicks may be the same as the grandparents of Catherine (or Kate) Dicks. If  I have it right, that also means that my wife’s step great grandmother nee Shave was also her 2nd cousin 3 times removed.

The Chromosome Browser

Here are the places where Esther matches her (half) niece and grand niece (my mother in law and wife). These matches represent the DNA From Esther’s father as that is the person these 3 women have in common as an ancestor.

Esther Chromosome Browser

Esther and my mother in law match on every chromosome except for #22 which is the shortest Chromosome. I am also interested in the X Chromosome. This is because these matches will be very specific. Let’s look back at the diagram with the purple squares and circles. These are the only places where Esther and Joan can share the X in a normal situation. Then Fred Upshall doesn’t get any X Chromosome from his father so the X matches must be from Kate Dix, born around 1851.

Unfortunately, this is not a normal situation. Here we have Esther who has parents that are related. I think that the relation is through the Dicks family. That means that Joan could also be matching Esther on Esther’s mother’s side. That would be Margaret Shave, up through her mother Melinda Kirby and Melinda’s mother Jane Dicks. Isn’t that confusing. However, as the most recent common ancestor between Esther and Joan is Esther’s father Fred, the matches most likely should be from him.

It actually gets even more complicated. Christopher Dicks didn’t get any X Chromosome from his father. However, Jane Dicks did. The unknown mother of the 2 sent her X Chromosome down to her 2 children. If any of those X Chromosome segments came down to my mother in law, she could have segments in common with both sides of her Aunt Esther’s parents. However, this would not be Dicks’ X Chromosome but the wife of a Dicks. Joan’s X matches on the Dicks’ side can only go back as far as Christopher Dicks.

X Triangulation

Triangulation is trying to figure out a common ancestor between using the matches of 3 people or more. The 3 people have to match on the same segment and match each other. What I did was choose Esther’s top 8 X Chromosome matches.

Esther X Chromosome Browse

Joan is #1. My wife is #3. My wife isn’t supposed to have a match that her mother doesn’t have, so I’ll disregard the blue match. It looks like the clusters could triangulate, but they don’t. Here is the X Matrix.

X Matrix Esther

This time my wife is last on the list. Note that Marie and her mom Joan only match each other and don’t match Esther’s other 6 top X Chromosome matches. That means there is no sense trying to triangulate these. These segments are likely to be from Esther’s mother’s side. That is, except for match #8 in the previous figure. That is Molly and her match doesn’t overlap with Joan’s or Marie’s, so we can’t tell which side of Esther she is matching (maternal or paternal/Shave or Upshall). It makes sense that Joan and Marie don’t match the other people. As I showed earlier, Joan and Marie’s X Chromosome match on the Dicks’ side ends with Chris Dicks. However, Esther’s X Chromosome match goes back a generation earlier, so she has many more chances for matches.

Autosomal DNA Matches: Not All One Way Or the Other

In looking at autosomal matches or triangulation groups, it is important to make sure that the matches are either on the maternal side or paternal side. However, what if the parents of your match (in this case Esther) had parents that were related to each other? It is not so clean cut.

Here are some of Esther’s top matches compared to Joan and each other:

Esther top matches matrix

Note that Joan has a larger match with Wallace than Esther does. Perhaps Wallace matches Joan on her father’s (non Upshall) side in addition to the Upshall side. Also Joan has no match with Nat. Either Joan didn’t get any of the DNA that Nat did or more likely this match is on one of Esther’s mother’s non-Dicks lines. Also some of these people may be matching Esther on her maternal and paternal sides where Joan only matches her on her Paternal side.

Just for fun, I checked if anyone else on the list had parents that were related. It turns out that Michael did. But not to the extent that Esther was. His common ancestor was about 5.6 generations back.

Effects of Endogamy On DNA Matches

Jim Bartlett explains the effect of relatives that marry in this Blog. When Esther’s parents married as the descendants of the same ancestor, they theoretically doubled their match with Joan. This effect multiplies when other matches also have the same ancestor more than once.  Joan and Esther might expect an average match amount of 850 cM as half aunt/neice. Their actual match total is 1090 cM. Also at the above chart, see how much more Esther matches people than Joan. I believe that this could be due to Esther’s multiple Dicks ancestors. This is true except for the match with Donald. Perhaps Donald is not related to the Dicks. When I check his ancestry, I don’t see any familiar surnames. Also no ancestry is mentioned in Newfoundland. So, maybe something to consider.

In Summary

  • My contact with Esther’s matches have resulted in good leads with people who have Newfoundland ancestry
  • Esther has, as expected, parents that are related
  • This relation appears to be on the Dicks side, based on both genealogy and surname ancestors of DNA matches
  • Further research should lead to linking up both sides of these Dicks family to a common set of ancestors.
  • Due to the irregular inheritance pattern of the X Chromosome, not many common matches were found
  • Endogamy results in more [presumed Dicks descendant] matches for Esther. This is compounded if her matches also have the presumed Dicks ancestor more than once.

 

Moving the Frazers Down the YDNA Tree

We have new YDNA results in for Jonathan and Paul. That’s good news. As you may recall, Jonathan had his YDNA tested about a year ago. He represents the James Line of Frazers. Then this year, Paul from the Archibald Line of Frazers tested. The tests were for 37 markers. The new tests are for 67 markers. Here is a tree that I sent to my cousin Paul who is not on the internet. Archibald, born around 1690 is believed to be our common ancestor and the husband of Mary Frazer at the top. Paul and Jonathan are 6th cousins, once removed based on our research. Paul is 2 generations below Hubert Frazer on the Archibald Branch and Jonathan is one generation below Walter Frazer.

YDNA Arch James Tree

Some Unexpected Results

  • Jonathan and Paul both showed a type of YDNA called R1a. I expected they would be R1b which is one of the most common Haplogroups in Europe. R1b is especially prevalent the further Northwest one travels in Europe.
  • Family Tree DNA (FTDNA) showed 3 mutations between Jonathan and Paul. I was expecting about zero to one. It turns out that all their mutations were on relatively fast moving markers.
  • Based on the markers, FTDNA puts people in a rough Haplogroup. Jonathan was put in R-M458 and Paul in R-M198 which is an even more broad or general category. With the new results, FTDNA has apparently backtracked and put Jonathan back into the more broad R198 Haplogroup. Usually, with more STR testing the Haplogroup should be more refined, not less.
  • At the different levels that FTDNA looks at (12, 25, 37 and 67 STRs), Paul matches on 4 people each. Normally there are many matches at the 12 level and the matches drop down to the 67 level. The apparent answer for this is that Paul has had more than the expected mutations in the earlier testing compared to Jonathan.

Genetic Distance

The Genetic Distance (GD) is the measure of how many differences there are in the STRs of 2 people. In the case of Jonathan and Paul, the GD was 3 for the 37 STRs and also 3 for the recent 67 marker test. That means all the differences were in the first 37 markers. Here are Jonathan’s results for the 37 STR test. The results of this test are also called a Haplotype.

Jonathan's 37 STRs

Here is what Paul has for STR results with the differences from Jonathan highlighted.

Paul's 67 STRs

Note that the there were 2 changes in the CDY marker. FTDNA informs me that they count this as one change as the markers represent a relatively fast mutating section of the YDNA. So in the roughly 260 years or 7 or 8 generations, there have been 4 mutations or a GD of 3 between Jonathan and Paul, assuming our genealogy is correct.

Refined TIP Report

FTDNA has a TIP Report that estimates the relationship likelihood of 2 YDNA matching people. For the previous 37 STR marker test, FTDNA thought that there was about a 44% chance that Jonathan and Paul were related at 8 generations apart. Now with the 67 marker test, that has gone up to about 65%. The percentage went up, because the GD was the same for 67 markers as it was for 37 markers. So it is more likely that these 2 are closely related. It is all based on statistics and probability.

Jonathan Paul TIP Report

Haplotypes and Haplogroups

The STR signature for Jonathan and Paul now consists of 67 markers. The combination of these markers is called a Haplotype. A Haplogroup is based on SNPs and is found one of 2 ways. The most accurate is by testing of the SNPs. The other way to estimate a Haplogroup is by the Haplotype. Jonathan and Paul have not had their SNPs tested, but have their STRs tested resulting in a Haplotype. Based on these STRs, people who are experts in looking at results can tell what your Haplogroup likely is. In our case, the L664 administer for the R1a project knew that if a DYS338 was 10, then it was veritably inevitable that if the SNP test for L664 was taken, then the tester would be positive for that SNP.

Climbing Down the YDNA Tree

FTDNA has Jonathan as R-M198 Haplogroup. This was from 6,500 B.C. Not good. Our astute L664 Dutch Administrator Martin got us down to 3,000 B.C. by noting that the Frazers are in the L664 Haplogroup. We appreciate him getting us an extra 3,500 years, but that still leaves us quite a way back in time. In my previous blog, Martin at first thought that the Frazers would not be in a SNP called S3477. Subsequently, Martin reasoned that we may be S3477 based on some similarities that he saw in the location of the Frazers and Prendergasts in Ireland. I made a prediction that the Frazers would be negative for S3477. The proof would be in the 67 STR test. If the Frazers did not have a value of 13 at DYS617, they would not belong to subgroup S3477. Let’s look at those results.

DYS617

It looks like I was right this time. Put another win in my win-loss column. The Frazers are not S3477. Speaking of SNPs, FTDNA recently came out with a new R1a panel for testing.

R1a Panel

All the grey hi-lighted SNPs above would apply to L664 Frazers. FTDNA boasts of over 40 L664 SNPs that they test for just under L664. This is a good introductory deal for $99 as it costs $39 to test a single SNP at FTDNA. To put these SNPs into context, here is how they look below our L664 Frazer Group.

L664 SNPs

The way it works, FTDNA doesn’t have to test 40 SNPs. For example, once they test S3477 and find it to be negative, they would not need to test the 10 SNPs below it. Remember, we were told that if the DYS617 STR marker was not 13, then we would be negative for S3477. In my previous blog, I mentioned that the L664 administrator didn’t think we belonged to the popular YP282 SNP. If that were true, then that would eliminate 13 SNPs. Likewise, Martin didn’t think we were part of the YP358 Haplogroup. It would be nice to know which branch the Frazers are on.

YDNA Matches

I had mentioned in an earlier blog that Paul had 4 matches at all of his levels of testing. This is quite unusual. Usually people have a large number of matches at the lower level of STR testing and fewer at the higher level. Apparently all of Paul’s mutations happened at these lower level of STR testing and wasn’t spread out over the 67 STRs. Here are his matches:

Paul's 67 STR Matches

These 4 matches are different than all the other levels of STR matches. At this level, Mr Frizelle drops out. This is not because he is not a match, but because Mr. Frizelle only tested up to the 37 STR level. Mr Latham was in the same category.  Replacing those 2 are a Stuart and a Grant.  However, the GD for these 2 are quite high and the relationship could go back to before the use of surnames. Jonathan’s matches appear to be with the same people that he matched at the 37 STR minus those who didn’t test at the 67 STR level.

Jonathan's 67 STR Matches

Here we see all of Paul’s YDNA relatives are on Jonathan’s list. So the YDNA relatives are starting to converge at this level – give or take 300 years! The Grants seem to be the most common name. It is possible that all these people came from the same area of Scotland and were related many years ago.

Summary and Future Considerations

  • A Genetic Distance of 3 at 67 STRs is closer than a match of 3 at 37 STRs for Jonathan and Paul. This was expected and supports the assumptions of our Frazer genealogy.
  • We are no further down the YDNA tree than L664 at this point. We know which part of the tree we are not on (S3477). To get further down the tree will take further analysis of the  recent 67 STR test or additional SNP testing.
  • We may want to look into the SNP panel for Jonathan and/or Paul to see where they are further down on the YDNA tree. I would assume that they both would have the same terminal SNP, so only one person would need to test for the panel of SNPs and the second could verify the terminal SNP with a single SNP test.
  • I will check with the L664 Administrator to see if he has any other analysis of the 67 STR results that would fine tune our Frazers’ place in the R1a Project