Using Visual Phasing

I did some visual phasing for Jane and her two siblings. Now Jane wants to see how the visual phasing can be used.

Mr Gray’s 8.5 cM Match

Jane’s first question came from a Mr Gray. He matched Jane and her two siblings at about 8.5 cM. Jane had this conundrum:

For me the match points to Beckham, for Heather it points to Beckham but for Alex it points to Hamilton. Unless it is on Mum’s side as indicated by Jackson for all three of us, but then why wouldn’t Mum share DNA with Mr. Gray?

Here is the right side of Jane’s Chromosome 2 Map. Jane is in the middle bar and her maternal side is on the top bar.

Here is where Jane matches Mr. Gray:

The answer for this one is that this is not a real match. Jane’s mother has tested and Jane has two sets of phased results based on her mother’s results. I ran Mr. Gray against Jane’s paternal and maternal kit and got no results. This means that the match is not real. Different analyses have been done for the chances of a certain size match being real or not. Here is one from Roberta Estes:

These studies seem to show that there is about a 50% chance of a 7 cM match being real. This goes up to a 66% chance of being real for an 8.5 cM match. It looks like Mr Gray fell into that 34% range. The only other possibility is that Jane’s mother could have had a false negative match with Mr. Gray. All in all, I would say that these results would be inconclusive. For me, I tend to say if there is no match on the two phased kits, then there is no match.

Second Example: An Adoptee at 23andme

Only Jane and her mom tested at 23andme. Here is how the adoptee matched Jane:

Chromosome 1:     108 to 151     26 cM

Chromosome 2:     225 to 230     8 cM

Chromosome 3:     104 to 145     41 cM

Chromosome 17     5.7 to 9.3      9 cM

By looking at the Visual Phasing Spreadsheet, Jane deduced that this was on her Beckham Line:

Jane is in the middle bar. The bottom of that bar is her paternal side where Bekham is in blue. I assume that Jane’s mother did not match this adoptee at 23andme.

I have Jane’s Chromosome 2 above where she matches Beckham between 225 and 230 on her paternal side.

The match is significant for Jane as the  match between 104 and 155 goes through her maternal crossover at position 110M. That would confirm that this is a paternal side match on Beckham.

Here is Chromosome 17:

It looks like I had trouble identifying the maternal grandparents, but Jane matches this adoptee in her paternal Beckham region. So I would agree with Jane that this match is on her Beckham side. It would help if this adoptee uploads his/her DNA results to gedmatch.com.

Peggy’s Maternal Match on Chromosome 13

Jane reports these results for Peggy’s matches to her and her siblings:

Heather     74 to 103     31.5 cM

Jane          74 to 91       14 cM

Alex – no DNA

Without looking at Jane’s map, it appears that Jane should have a maternal crossover at around 91M.

Jane has a maternal crossover at 90 on the map where her DNA goes from Adair to Jackson. Alex is on the top row. He has a large maternal Jackson segment which explains why he has no match with Peggy. Jane has determined correctly that this match is along the line of her maternal Adair grandmother.

A Paternal Hillock Match on Chromosome 2

This match has a similar pattern to the previous one. Here Hillock matches:

Jane     40 to 79         41 cM

Alex      119 to 208      8.4 cM

Heather – no DNA

Two siblings match Hillock and one does not. However, in this case, the two siblings match at different locations on Chromosome 2.

Jane again is shown on the middle bar. The paternal side is on the bottom where she matches on her Hamilton grandfather. The match with Alex on the top shows that he should be matching on the Beckham paternal grandmother. However, this is a small match. It is time to check Alex’ paternal phased kit against Hillock.

I forgot that I had lowered the threshold and was surprised to come up with this result:

The location of this match is within the orange Hamilton segment for Alex’ top bar. The match that Jane reported for Alex should have been 199-208 (unphased). That means that my red circle above on Alex’ bar above is not correct.

Summary and Conclusions

  • Jane is fortunate to have a mother whose DNA could be tested. This gives her paternal and maternal matches for Jane and her two siblings.
  • It is important to know first whether the match is on the maternal or paternal side.
  • If there is no maternal or paternal match, I consider that to not be a match.
  • Jane’s conclusions appear to be correct as far as on which grandparent line her matches are on. However, it is important to make sure that the position numbers of the matches are correct to match up with the right grandparent.

 

Visual Phasing of My Canadian Frazer Relatives

One of the Canadian Line of Frazers descends from my 2nd great-grandfather’s brother. His name was Richard Frazer born 1830 in Roscommon County, Ireland:

My 2nd great-grandfather was George William Frazer, born about 1838. Here are the two lines as shown with their descendants that have taken a DNA test:

Visual Phasing

Visual Phasing is comparing the DNA results of three siblings (or more) and figuring out from which of their four grandparents they got their DNA from on each of their chromosomes. I did that with myself and my four tested siblings. Now on the yellow line we have Susan, Doreen and Ken who have all had their DNA tested. I am 4th cousins to Susan, Doreen and Ken. However, by figuring out where we got our DNA from, it will be almost like comparing our grandparents to each other. Richard Price “Pat” Frazer will be compared with my grandmother, Marion Margaret Hartley. These grandparents would have been second cousins to each other. Now we didn’t get all our grandparents’ DNA, but what we did get, we will be able to identify and compare.

Steven Fox’s Excel Spreadsheet for Visual Phasing.

If you do a Google search for Steven Fox and Visual Phasing, you will find a lot of information and blogs of other’s that have used his spreadsheet. It is possible to do Visual Phasing without his spreadsheet, but his spreadsheet makes it easier, more organized and more standardized.

Jumping in with Chromosome 22

Chromosome 22 is the shortest, so some people start with that one. There should be fewer crossovers on this Chromosome. The downside is that there are also usually fewer cousin matches on this short Chromosome.

This image is small but if you click on it, it should get larger. The top bar compares Ken (K) and Susan (S). The vertical lines represent crossovers where DNA received changes from one grandparent to another. The location of these changes or crossovers are important. It is also important to know to which sibling these crossovers get assigned to. The first crossover is found where the match between Ken and Susan goes from HIR or FIR. HIR is Half Identical Region. That means that Ken and Susan both got the DNA from one of their 4 grandparents at that segment. The problem is, we don’t know which grandparent. After the crossover, represented by the first vertical line, Ken and Susan have an FIR. This is a Fully Identical Region. That means that Ken and Susan both got their DNA in that segment from two of the same grandparents. One matching grandparent was on the maternal side and one was on the paternal side.

Finding the Location for the First Crossover

The second crossover is easy to find, because it is at the same place where Susan and Doreen go from no match to a match. This spot is recorded by Gedmatch at 23,564,890. I could call that 24M or 23.6 depending on how exact I want to be. In the case above, I called it 23.6. To get the first crossover, we have to look at the comparison between Ken and Susan at full resolution at Gedmatch.

Here is the first part of Ken and Susan’s comparison at Chromosome 22:

This says that the beginning of Chromosome 22 doesn’t even start until position 14.5M where M is million. Every little ^ is one million. So counting back from 20M, Ken and Susan go from HIR to FIR between 15 and 16M. I’ll call it 15.7M. The pink part is a centromere. Usually this would be at the center of the chromosome, but there must be information at the very start of Chromosome 22 that isn’t used.

One Crossover or Two on the Right Side of Chr22?

Here is a blowup of the last two crossovers. On the top bar, there is a spot where Ken and Susan go from HIR to FIR. However, this does not line up with position 45.4M where Ken and Doreen go from no match to an FIR. However, to be sure, I need to know where Ken and Susan go from an HIR to an FIR.

According to Gedmatch, the end of Ken and Susan’s match is at 49.5M. That means the last ^ is 49M. That means that Ken and Susan’s change from HIR to FIR is between 45 and 46. I’ll say 45.7. This is pretty close to 45.4, so a judgement call. I’ll just leave it as I had it.

Assigning the Crossovers

The crossover usually gets assigned to the person that is in two of the changes. This is easier to see in the second crossover. There is a change in the top and bottom comparisons. The first comparison is Ken and Susan. The last comparison is between Susan and Doreen. Susan is the common denominator, so she gets the crossover.

I gave the first crossover to Ken:

I had forgotten why. Now I remember. I think that there is actually a very small match that doesn’t show up between Ken and Doreen at the beginning of the Chromosome. It is very subtle, but I’d say that there is more green in the beginning of the K&D comparison compared to the S&D comparison. These crossover assignments can be a bit of an art. If I had lowered the match resolution it should have shown up as a blue match there.

The last two crossovers were not easy either:

These changes have to come in pairs. That means that I had to treat the little segment between D and S as an FIR on the bottom comparison. That means that there are two changes for Doreen first making that her crossover. Then there are two changes for Susan (top and bottom comparisons).

Working on the Segments

This image is the end result, but I will say how I got there. First I started with K&S. They have a FIR for the second segment. This shows as dark green. In the image above, that means that Ken and Susan have two of the same colors. They match with one maternal grandparent and one paternal grandparent. Because Ken has no crossover to the right of the second segment, I can extend that DNA all the way to the right end of the Chromosome on both Ken’s maternal and paternal sides. Also in the second segment, Doreen has no matches with Ken or Susan. That means she has the spouse of the other two grandparents in that slot. So instead of orange and purple, she gets blue and green there. She has no crossover to the left and none on the right until near the end of the Chromosome.

At some point I have to deal with a HIR. I did that with Susan after the second segment. Susan has a HIR to Ken and Doreen. I made Susan match Ken’s purple but not his orange. We already knew that Ken and Doreen had no match in the third segment, so Doreen got the opposite colors there.

The last three bars in the image above represent Gladys’ match with Ken, Susan and Doreen. She shows a match with only Ken and not Doreen nor Susan. This has to be in Ken’s orange section as that is the only place along the match with Gladys that one of his segments is different than Doreen’s or Susan’s.

Gladys and Ken only match on the Frazer Line. Gladys has no Gray DNA. That means orange has to be Frazer. The only other paternal side is Gray so blue has to be Gray. I don’t have specific information on Ken, Susan and Doreen’s maternal side, so those are just labeled G3 and G4 for now. Actually, they should be M1 and M2 for maternal grandparents:

Ken and siblings’ father’s name is Stefansson and their mother’s surname is Gudmundsdottir.

How to Use Visual Phasing

Now that we know where Ken, Susan and Doreen got their DNA from on most of their Chromosome 22, what can we do with this information?

Part of the information is educational. Knowing how our DNA recombined by way of our grandparents is interesting. It is a kind of snapshot of what went into our makeup at conception.

Another part has to do with DNA matches. This focus us as to where our matches are. However, there is still one hitch. We still need to know if our matches are on our maternal or paternal side. If we don’t have a parent to compare our matches with, then it is possible to compare matches with known relatives.

Use of Crossovers

The exception to this is if Susan were to have a match that started before her crossover at 23.6 and continued beyond it. For a match that goes through a crossover, it has to be on the other side. So for such a match for Susan, this could not be a Frazer match but must be on her mother’s side.

These visual phasing maps work best when you have downloaded all your matches. Then you will know for every Chromosome which grandparent they will match depending where in the Chromosome your match is. If you have a lot of matches that end at a certain place and then other matches that start up again at that spot, that could be indicative of a crossover.

Separating Real Matches from Far Away or False Matches

Another way to use these results is to tell if a match is real or not. Here is part of mapped Chromosome 22 showing a small match between Doreen and Michael of the Frazer Project:

It shows that Michael only matches Doreen and not Susan nor Ken. Michael’s match is indicated by a blue bar at the bottome right of the image above. However, Doreen has no Frazer DNA in that segment. She only has Gray DNA on her paternal side. So, it may mean that Doreen matches Michael on the Gray line going way back or less likely on the Icelandic Line. But Michael could not be matching on the Frazer Line assuming I have mapped this correctly. As I mentioned, this is a small match of 5.6 cM. Matches under 7 cM have more than a 50% chance of not being real matches.

Comparison with My Visual Phasing

Here is my Chromosome 22. It is mapped along with three of my siblings. I didn’t get around to mapping Lori.

When I compare the two maps, I can see where my siblings have the potential to match with Ken, Doreen and Susan along the Frazer Line. It doesn’t mean we have to match there. For example, my Frazer grandparent DNA also has Clarke and McMaster DNA that Ken, Doreen and Susan don’t have. Likewise, Ken, Susan and Doreen have other DNA in their Frazer line that I would not share.

Extra Chromosome Mapped

While I was at it, I mapped Chromosome 12:

This was a little easier, because more cousins matched on this Chromosome. One interesting thing about this Chromosome is that between about 88 and 104M, there is no Frazer DNA. That means that if any of these siblings have a match in that area, it could not be a Frazer match.

This map had three spots that appeared to go from a no-match to a FIR or from a FIR to a no-match. This is impossible, so there needs to be a HIR transition. That is why there are three places where the crossovers are close.

Summary and Conclusions

  • Visual Mapping shows the process of our makeup based on how our grandparents’ DNA combined to form us
  • This mapping can be helpful in identifying DNA matches
  • Special attention should be made to crossovers. That is where DNA on one side of our Chromosome changed from one grandparent to the other.
  • If a match goes through a maternal crossover, for example, it means that match must be on the paternal side.
  • A lot of genetic genealogy is about separating out the DNA and visual phasing goes a long way in doing this.
  • Other sorting can be done by names or location of matches. As Ken, Doreen and Susan’s mom is Icelandic a clue as to the matches’ names, locations or genealogy can be a hint as far as placing them.
  • Visual Phasing works best when there are good cousin matches on all four grandparent sides.

 

 

 

 

 

Two Person Hartley Visual Phasing

I’ve had a FTDNA kit hanging around for my father’s elderly cousin. I’ve had it since last Summer, but haven’t gotten in touch with my second cousin Lisa to see if I could get her uncle tested. This would be important, because I have test results for Lisa’s dad Jim and her Aunt Joyce. The third sibling Ralph would make it easier to perform Visual Phasing.

Visual Phasing

Visual Phasing is comparing siblings’ DNA results in a Chromosome Browser. By looking at changes and comparisons in the Browser as well as matches to known cousins, it is possible to find out what portions of the siblings’ DNA came from which grandparent. For me, this is important as I am interested in separating out matches between my great grandparents Hartley and Snell. Jim and Joyce’s maternal grandparents were James Hartley and Annie Snell. Annie’s ancestors went back to SE Massachusetts Colonial times. James ancestors were from NE Lancashire. I’m stuck on Hartley genealogy in Trawden, Lancashire around 1800. This is due to the fact that there were too many Hartleys in the area at the time to tell one from another based on vital records. Finding Lancashire Hartley ancestor DNA matches may help me break down my Hartley genealogical brick wall.

Joyce and Jim’s Genealogy

The goal of visual phasing is to figure out what parts of Gurney, Rounesville, Harltey and Snell contributed to Jim and Joyce’s DNA. In doing this, it would help to have matches from fairly close (but not too close) relatives on all four lines.

Comparing Jim to Joyce on Chromosome 11

I’ll just jump in and start with Chromosome 11. This is midway between 1 and 22. Here is the comparison between Jim and Joyce:

  • The blue line is where Jim and Joyce match each other
  • Within the blue line there are two types of matches
  • The yellow area is a single match. This is also called a Half Identical Region (HIR). This means that Joyce and Jim get their DNA from one shared grandparent A, but don’t match on grandparent B, C or D. We don’t know now if granparent A is on the maternal or paternal side.
  • The green is a double match. That is called a Fully Identifal Region or FIR. In that area they got the same DNA on their maternal and paternal side of Chromosome 11. That also means that they share the DNA from the same maternal grandparent and the same paternal grandparent
  • The grey, non-blue area (below) and the red area above is where Joyce and Jim do not match. That means that Joyce gets DNA from Maternal grandparent A and Paternal grandparent C while Jim gets his DNA in that area from Maternal grandparent B and Paternal grandparent D
  • At each vertical line above, there is a crossover where Jim or Joyce’s DNA goes from one grandparent to another.

Let’s Start Two Person Visual Phasing

Here is a start. In about the middle of the Chromosome there is a green FIR. That means that Jim and Joyce got their DNA from the same maternal and paternal grandparents. Those grandparents are represented by blue and orange segments. There are crossovers on the right and left of these segments, but we don’t know if the crossovers are for Jim or Joyce (or one for Joyce and one for Jim).

It would be nice to know where the changes take place, so I go to gedmatch.com for that. At gedmatch I compare Joyce to Jim in the chromosome browser at full resolution.

The pink area is the centromere of Chromosome 11. Every ^ is one million places. The start of the green HIR counting back from 60M is 57M.

Here I added the 57 before ‘Chromosome 11’ above. I also added some other crossover locations.

Cousin Matches

I am stuck already in my analysis, so I need some cousin matches. These would ideally be at the level of second cousin matches. At the level of second cousin, you match on only one grandparent. Most known matches matches share Hartley and Snell grandparents, so that is a problem.

Shared Ancestor Hints (SAHs) at AncestryDNA

Joyce’s results are at AncestryDNA. There, she has Shared Ancestor Hints. Those Hints are where Joyce has a family tree match and a tree match. Here is an SAH that Joyce has with Chuck:

Chuck is at the perfect level as he is a 2nd cousin. However, he has not uploaded his DNA to gedmatch for comparison. Ancestry does not show on what Chromosomes you match, so that is a problem. We need chromosome match information for DNA mapping.

Back to Gedmatch

Because many at AncestryDNA don’t upload to Gedmatch, I’ll go back to Gedmatch and look for matches there.

Here is a very interesting match that Sumner and Heather have with Joyce at Gedmatch. This shows that Joyce has an estimated by DNA common ancestor between 3.7 and 3.9 generations away. They also share autosomal DNA and X Chromosome DNA. These two are also at Ancestry and show up on Joyce’s Shared Ancestor Hints.

Here, Joyce and Sumner are 4th cousins by shared trees. However, note that this is only hint 1 of 3. HInt two also goes back to Joyce’s Rounseville grandparent at firth cousin twice removed. Here is Hint 3:

This Snell connection is at 7th cousin once removed. There has to be a very low chance of a DNA match that far out – especially compared to a 4th cousin match . However, this is interesting as it shows that Joyce has two paternal matches with this person and one more distant maternal match.

Here are the important details of the match between Joyce and Sumner:

This shows that Joyce and Sumner match on four different chromosomes, but not Chromosome 11. OK, back to the drawing board. I’ll start over with Chromosome 7. Sumner and Joyce have a pretty good match there.

Chromosome 7 Visual Phase Two Person Map

Note that Joyce’s Chromosome 7 match is from 149 to 158M. That is at the right side of Chromosome 7. It is possible that the 149M could mark Joyce’s paternal crossover. I am going to start from the right of the Chromosome and give Jim and Joyce four different colors there. This will represent all four of their grandparents. I can do that because Jim and Joyce don’t match each other at all in that segment.

Here I have put Joyce in for a possible to likely crossover at 149. Remember that Joyce and Jim don’t match each other at all after 149M. That means that Jim won’t match Sumner either. I checked gedmatch and he doesn’t as expected. Next, I’ll assign Sumner’s match to Joyce on either her green or brown side. I’ll randomly choose green. That puts the paternal side on the top for Jim and Joyce:

Becuase Joyce’s green paternal segment is Rounesville, that means that Jim’s orange segment must be the paternal husband, Gurney.

Next, I would like to check the paternal crossover for Joyce. The recommendation at the Facebook Visual Phasing side is to look for ‘stranger matches’.

Stranger Matches

If I see that Jim has a match or matches that go across the 149 crossover line, then I can assume that he has no crossover there. The hitch is that the match going over the 149 line needs to be on Jim’s paternal side on the top of his Chromosome 7.

Here is a spreadsheet of Jim’s matches on Chromosome 7. Jim’s match with Tim goes clearly from 138-155M. That meets one requirement. Is this a paternal or maternal match for Jim? My thought was that if this match was maternal, then Tim should match my sister Heidi at the top and me at the bottom of the list in blue. I checked and Tim only matched Jim. That means that the crossover belongs to Joyce and is likely on her paternal side. The only thing I didn’t rule out is that the crossover could possibly be on Joyce’s maternal side.

Here I went with my original guess that Joyce’s crossover was on her paternal Gurney/Rounesville side. Because I gave the crossover to Joyce’s paternal side, that meant that there was no other crossover at 149 and I moved the maternal segments to the left. I still have figured out whether Hartley or Snell is blue or brown. Next note that the segment from 110 to 126M is a no-match segment. That means that there must be a maternal crossover next. The reason for that is that no-match means four different colors. Jim and Joyce already have different colors on the maternal side. If we change one of those colors with a maternal crossover, there will be a match between 110 and 126M.

In order to get a no-match from 110-126M, Jim or Joyce’s DNA must be Rounseville from 110 to 126M.

Stranger Match or More Cousin Matches?

I really should go with both, but I’ll start with the stranger match. Jim has matches between 105 and 134 showing no crossover there. When I look at one of those matches and run those that are in common, I get this:

#1 is Jim’s sister Joyce. 2-6 are the strangers and #7 is actually a 2nd cousi of mine, but it could be from a match on another line. So Jim is matching the strangers in that 105 to 134M area. However, he is matching Joyce starting at 126. That gives me the impression that it is Joyce that has the crossover. On the other hand, I don’t see any of Joyce’s matches on her match list that go through 126M.

i am moving slowly from right to left on Chromosome 7. The segments that I am really interested in, I have no information on – except that one is Snell and one is Hartley and they appear to be relatively large segments, so far.

Phasing by Geography

I had mentioned that Snell’s ancestors were from SE Massachusetts going way back. The Hartleys came to the US from Lancashire in the last half of the 1800’s. As far as I know, the Gurneys and Rounesvilles have been around SE Massachusetts for several hundreds of years also. When I look at Joyce’s matches at Chromosome 7, I see some interesting emails. Between 155M and the end of Chromosome 7, Joyce has three small matches with people three people that have nz, au or uk in their email addresses. That gives me the opinion that at least from 154M to the Joyce could have Hartley DNA. That also brings up the question as to whether Joyce has a maternal or paternal crossover at 149M. If I go with what we had already, I would get this:

Starting to Visually Phase Chromosome 8

I can come back to Chromosome 7 at some time. I’m looking at Chromosome 8 as I wrote a Blog about a Lancashire matcher here. Here is how Anne matched Joyce and two of my second cousins:

The important part is that Anne matches Joyce from about 17 to 59M. That is a pretty good match. Here is the common ancestor:

The other important thing is that even though the match points back to Howorth, this is on Joyce and Jim’s Hartley grandparent line.

Here is how Jim and Joyce match each other:

Here I did something different. I started by mapping a HIR or Half Identical Region. That means that one grandparent matched and the other two did not. We know that Joyce matched on the Hartley segment and Jim did not.

That means that the maternal Hartley/Snell side is on the bottom of their Chromosome 8. From here, we can logic a few more segments. Going from HIR to the no-match left, that means the top part will have to change for there to be no match at the beginning of Chromosome 8. Using similar logic, for all to match (in the HIR region), the crossover will have to be on the bottom of Chromosome 8.

Next, on Joyce’s match list, I picked someone who she matched that went through the 70.7M crossover.

I picked the 18 cM match. Then I picked people that matched both Joyce and the stranger’s 18 cM match.

#1 is Jocye’s match to her brother Jim. The next three matches go up to 74, so they go through the crossover. #5 is our stranger, Sheila with the 18 cM match. There is another interesting thing about Match #2. That is Jo who is on Ancestry with a private tree. However, when I click on her name, it says she is from Lancashire, England. Someone with a tree at Ancestry and DNA at gedmatch is good news to me, so I wrote an email to her.

Here is another piece of the puzzle:

I don’t know what the orange represents, but I don’t match Joyce and Jim on that side, so it isn’t as important to me. I was interested in separating the green DNA from the blue – or the Snell from the Harltey DNA. I was able to do that thanks to visual phasing and a match with Anne.

Wrapping It Up

  • It is possible to do some visual phasing with only two siblings. However, cousin matches, and stranger matches are needed.
  • Geographical phasing is also important. I like the use of email extensions to identify non-US matches.
  • Mapping my father’s two cousins is important in separating my Lanchashire ancestors from my colonial Massachusetts ancestors.
  • Work is needed to get AncestryDNA testers to upload their results to gedmatch.com
  • More matches could be found by checking FTDNA
  • More work is needed in tracking down genealogies of gedmatch mathes. This would help identify segmens of visually mapped chromosomes.
  • Attention to mapped segments of interest (in this case Hartley) can lead to matches to follow-up with.

Chasing Down My Wife’s Rooney Connections

My wife’s father is half Irish and half French Canadian. On the French Canadian side there seems to be  a lot of genealogy and a lot of DNA matches. On the Irish side, there is a not so much genealogy and a lot less identified DNA matches.

Mapping the French Canadian and Irish In Laws

I have used a method to map out my father in law’s DNA that he got from his four grandparents. To do this, I compared him to his two sisters, Lorraine and Virginia. Here is their Chromosome 14.

The good news was that I could map the Chromosomes by looking at the DNA results of the three siblings compared to each other. Then I could find many matches to reference the French Canadian side. That got me the LeFevre and Pouliot grandparents above. The problem was that I couldn’t find enough matches to reference the Irish side.

Gaby to the rescue

However, on AncestryDNA I found my wife’s 2nd cousin on the Irish side. Because of Gaby, I can now tell which of my father in law’s grandparents are Irish.

Any DNA matches that Gaby has in common with Lorraine, Richard or Virginia are Irish. Gaby and my wife Marie, share the same Butler and Kerivan Irish ancestors. The next problem is that we can’t tell whether these matches are Kerivan or Butler.

Working Gedmatch To Get Kerivan and/or Butler Matches

In order to separate the Butlers from the Kerivans, we need to find matches that are further out. To find these I looked at DNA matches at Gedmatch that matched both Gaby and Lorraine. I used Lorraine because she was tested at AncestryDNA. The matches would be on the Irish side. That was the first cut. Next, I hoped that some of these matches would have trees at Ancestry that would match my in-law’s tree.

For example, here is someone that matched both Lorraine and Gaby on our example Chromosome 14.

The above image shows how Lorraine matches someone with a Rooney name (#1) and Gaby (#2). This tells me that this Rooney match is on the paternal side or Irish side, so that is also good. The other good thing is that my father in law’s grandmother’s mother was a Rooney:

All I have to show is that the match indicated in yellow above with the Rooney name is related to Alice Mary Rooney above. There were other common surnames, so the match didn’t have to be a Rooney. However, I noticed that there were some Rooneys in Massachusetts which is where my wife’s Rooney ancestors lived. Based on that, I thought that it would be a good idea to start with Rooney.

Doing the Rooney Genealogy

Lorraine’s Rooney AncestryDNA match that is also at Gedmatch and matches with Gaby at Chromosome has a Rooney Tree:

However, these two trees seem a little out of whack. Maybe Timothy Rooney in my wife’s tree could be a brother of Terrance Rooney in the Rooney tree?

A third Rooney Tree

I found another Rooney tree as an Ancestry Hint. It looks like this in a different view:

This tree shows that Timothy Rooney had two wives. It appears that Margaret Gorman was the first wife and had a John Rooney born 1827. Apparently Ann Nancy Lilley was the second wife and had Alice Mary Rooney. That could explain why the two trees didn’t match up. This tree shows the Terrence Rooney from the Rooney Tree as the same Timothy Rooney from my tree.

Putting the rooney trees together

Assuming that the Rooney Tree reconciliation was correct, the Rooney DNA match on the bottom right in purple would be a 1/2 third cousin once removed to my father in law Richard and his two sisters.

Back to the Chromosome 14 Map

That looks better. We now have the paternal side thanks to Gaby and a Rooney match. When I check the Rooney match, he matches Lorraine and Richard, but not Virginia.

The yellow matches on the Gedmatch Chromosome browser correspond with the green in the Chromosome 14 map above. The crossover for Richard at 54M also shows up.

The other good thing about the new Chromosome map is that it shows where the Butler matches would be. This is like a spy glass looking into the past. A match on the Butler side is like a match with Virginia’s grandfather who was born in 1875. Matches to these grandparents should be helpful in straightening out my wife’s Irish genealogy.

Summary

  • Use a paternal cousin to find other paternal cousin matches that are more distant
  • Connect that further out cousin to a known ancestor
  • Use that further out cousin match to complete a Chromosome map
  • Use that completed Chromosome map to identify other cousins as they match in identified areas of the Chromosome map representing grandparents of my father in law.
  • Use those identified matches to focus on further genealogy and break down former research barriers.
  • This method works best with people that have DNA testing results at both Gedmatch and Ancestry.

Visual Phasing Butler DNA on Chromosome 2

The Butlers are my wife’s family. Over a year ago, I wrote a Blog called “Uncle Naffy, DNA and the Butler Brick Wall“. In that Blog, I wrote about how a match with Uncle Naffy who is believed to be a Crowley relative helped in producing a breakthrough in the Butler genealogy. Uncle Naffy is a rare paternal Butler match. Most of the Butler matches have been maternal on the French Canadian side.

Visual Mapping of Chromosomes

Since writing the Uncle Naffy Blog, I have also become aware of a tool to map Chromosomes. This visual mapping procedure was developed by Kathy Johnston. As the Uncle Naffy match was on Chromosome 2, why not map that Chromosome? In order to map my father in law Richard’s four grandparents, I need his results and two siblings. Since my Uncle Naffy Blog, I have tested Richards two sisters: Lorraine and Virginia. When I compare these three siblings at their Chromosomes 2, this is what I get:

  • In the green regions, the paired-up siblings share the DNA from two of their same grandparents.
  • In the yellow regions, the sibling pairs share one grandparent
  • In the red regions, the sibling pairs share no grandparents in common. That means they have their DNA from the opposite grandparent pair.
  • The areas between the green, yellow and red regions with the vertical lines added are the crossovers.
  • The crossovers are assigned to the person who has the most shared crossover regions
  • The numbers added are the approximate positions in millions of the crossovers
assigning the crossovers

Lorraine gets the first crossover because she is the one in common in the first two comparisons where the match goes from green to yellow. The other crossovers are from the same reasoning. Richard is in the top and bottom comparisons.

Mapping three butlers

I start by using two colors where Richard and Virginia match representing their shared Fully Identical Region (FIR) shared in green. These two colors represent the same grandparents that Richard and Virginia inherited their DNA from – one on their maternal side and one on their paternal side.

Richard is stuck between his two crossover points (R) but Virginia can go out in either direction to her two crossover points (V):

By using other green areas or Fully Identical Regions (FIRs) and areas where pairs don’t match using opposite colors, I get this:

This leaves a few holes. At this point we need to select a Half Identical Region (HIR). It would be nice to get Lorraine to the right side as she doesn’t have any more crossovers there. Lorraine and Virginia share a HIR from 128 to 204, so we will pick one color from each on Lorraine’s row and extend those to the right. As I mentioned, she has no crossovers we know about there to stop her.

I can fill in a little more using the FIRs and no-match areas.

Now we have four relative grandparents without names in a lot of these three siblings chromosomes. Using known matches, we can fill some of these in. The paternal grandparents are Butler and Kerivan. The maternal grandparents are LeFevre and Pouliot.

Adding relatives

The best known relative for this purpose is a 2nd cousin. Richard and his sisters have two known second cousins on the Pouliot side:

Here, the Pouliot matches didn’t help, due to a blank space. The Uncle Naffy match, assumed to be a Butler match helped. There was only one place that it could go. That sets the paternal side and also will make the green be Kerivan.

What now?

At this point I have two options. One, I can look for more matches or I can try to re-do the mapping. I tried looking around Ancestry for more matches. There are plenty of Pouliot matches there, but it is difficult to trace them to Gedmatch, or perhaps the Pouliot matches are not uploaded to Gedmatch. Right now, we have a proposed match identifying the paternal side. It would be nice to somehow get both sides.

Second Try at mapping Chromosome 2

In our two reference matches, we have Richard. He matches Pouliot and he matches Uncle Naffy. Also Lorraine matches both those reference matches. So let’s work on our Lorraine/Richard matches, rather than concentrating on Virginia who didn’t have too many crossovers. Between 128 and 149 Lorraine and Richard don’t match. This will be represented by two opposite colors.

Next, expand the segments to the crossovers:

That is good because our segments are now over our reference matches (Richard is over the Pouliot match and Lorraine’s represented DNA is over her Butler match). Next we can use the relationships with Lorraine on the right to create new segments from their relative grandparents (no pun intended).

Now we have another problem, we need both Lorraine and Richard to be expanded to the blue and yellow matches. Perhaps if we extend Lorraine to the left with an HIR, then Virginia will be opposite of Lorraine and Richard oppose of Virginia, it will work out to fill in the segments over our reference relative matches.

There. Now all we have to do is match the Pouliot (blue) and the Butler (yellow). The only colors the same between Lorraine and Richard above the blue is green, so that has to be Pouliot which is maternal. That means that the maternal side Butler is now top bar. The only color on the top bar (or either bar, for that matter) that is the same over the yellow Butler match is purple.

But there is more that I can do. Notice on Virginia’s Chromosome. I haven’t moved her over to her left-most crossover. This should help fill in some more.

Note that two crossovers in a row in a HIR cause a problem such as the L-L on the left side and the V-V on the right. However, I’m happy with the results. I now have the first Chromosome with 4 Butler grandparents. This is based on the presumption that Uncle Naffy is a Crowley relative who is ancestral to the Butler side. Virginia will be a good person to look for Kerivan matches. Lorraine looks like the best shot for checking Butler matches on Chromosome 2.

Finding Crossovers by comparing first cousins

It appears that we can look at these three sibling’s maternal cousins (Pat and Joe) to determine more crossovers. Here are Lorraine’s matches:

The crossover between Lorraine and Patricia is not clear by looking at the first yellow bar. But look above. This bar has a break between 67.4 and 67.9M that is not visible. That tells me that the maternal crossover for Lorraine occurs at that location.

Other crossovers

For Lorraines’ first crossover position, I will need to look at the Gedmatch expanded view. When I compare Lorraine to Virginia, I choose the full resolution box and get this:

Each up arrow (^) is 1M, so Lorraine’s crossover is at 27M. A little further on the same comparison is the change from HIR to FIR:

I would estimate this crossover at about 36.7M. These are the numbers for Lorraine and Virginia’s first crossovers:

Note that Richard’s first crossover is very close to his sister Virginia’s. Here is a closeup view of Richard’s first crossover using his comparison to Virginia:

There is a ^ mark right in the middle of the HIR for Richard and Virginia. Counting back from 40, that mark is 37. The FIR starts up again about 37.5, so that will be Richard’s first crossover.

Richard’s relatives on the chromosome browser

Here is a comparison of Richard to a nephew, two maternal first cousins and two second cousins on the Pouliot side, John as a nephew, may match on the maternal or paternal side. He is the son of another sibling of this trio not tested. Here, he appears to match on Richard’s paternal side

I’ll add in that maternal crossover for Richard:

Then the HIRs are added in for Lorraine and Virginia:

Once Richard’s crossover was found to be on the maternal side, that required his sisters’ first crossovers to be on the maternal side also.

filling in virginia’s blank spot

We just have a little area to fill in past 200M for Virginia. Is her crossover paternal or maternal? Here is how Virginia matches her nephew, and two maternal first cousins. The numbers that we will be looking for will be 224 and 227.

What we see is 224M. That means to me that there is no maternal crossover at 227 as all the matches carry on to what looks to be the end of the Chromosome. Therefore the 227 crossover must be on the paternal (Butler/Kerivan) side. Here is the completed Chromosome 2 map.

One observation is that the trio of siblings comes up short on Butler DNA (purple) for about the first third of the Chromosome.

Here are my wife’s father’s four grandparents all born in the 1870’s:

Bonus Feature: My Wife’s DNA a la Blaine Bettinger

Blaine Bettinger recently wrote a great instructive 5 Part Blog on Visual Phasing. My Blogs are my muddling and meddling with DNA. Blaine’s Blogs on the other hand are instructive. Part Four of Blaine’s Series shows how to take the results of the parent (and Aunts in this case) and apply them to the child (in this case my wife). I’ll look at Blaine’s Part Four and apply it to my wife. Here is Marie’s Dad’s Chromosome 2:

This tells me where Marie may or may not be getting DNA. She will get half of her DNA from her dad, but that will be a full Chromosome. To the extent that she gets her dad’s paternal side what she gets will be only Kerivan in the first two thirds and Butler in the last third.

Marie compared to Aunts Lorraine and virginia

Here I copied Blaine’s format, but was tempted to add some vertical lines. The browser images compare Marie to Aunt Lorraine and her Aunt Virginia.

Marie has one empty bar which we hope to fill with her 4 paternal great grandparents. Comparing Marie to Lorraine, they share a segment and then they don’t. With Marie compared to Virginia, the two keep sharing the same segment apparently. This appears to be Pouliot as Lorraine has a crossover from Pouliot to LeFevre right where she stops matching Marie. One way to check this is by comparing Marie to her dad’s maternal cousin. Marie matches that cousin in this segment which agrees with my reasoning as Pouliot is a maternal match.

For the next segment, it appears I can use the same reasoning. Marie matches Lorraine and Virginia but this time Lorraine’s match drops off right where she has a crossover from Kerivan to Butler. That makes me think that the match there is with Kerivan. Another way to look at it is that it has to be Kerivan there as Lorraine and Virginia don’t share a common grandparent on their maternal side in that location. Marie has to have the maternal Kerivan DNA in that location.

The next segment has to be Kerivan or Pouliot. Marie matches Virginia there who has Kerivan DNA, so that has to be it. That extends Marie’s Kerivan DNA. Next is the largest segment. Marie matches neither of her Aunts in that segment. The only grandparent that her Aunts don’t match in that segment is Pouliot. So far, Marie has no Butler nor LeFevre DNA:

In the next to the last segment, Marie does not match Virginia. That leaves her with Butler or Pouliot DNA. However, that is not helpful as Marie gets Butler or LeFevre from her father. Marie matches Lorraine, but that also could be Butler or LeFevre. It’s a split decision. In the last segment, there is a clue. Virginia matches Marie for Virginia’s entire maternal LeFevre segment. So that has to be LeFevre. If it was a Butler match, it wouldn’t be the entire segment as Virginia has some Kerivan in there that Marie could not have inherited from her father in that location.

Let’s try to reason through the empty space again.

  • Lorraine – matches on either Butler or LeFevre
  • Virginia – doesn’t match on Kerivan nor Pouliot which leaves, again, Butler or LeFevre
  • Richard – has on Butler or LeFevre

I suppose that this segment would more likely be LeFevre than Butler as larger segments are the rule more than smaller ones between father and daughter, however, I have no certaintly with that, so I will leave the segment blank for now.

The maternal cousins to the rescue

It’s time to bring back Pat and Joe. They are Marie’s father’s maternal first cousins. Here is where Marie matches them:

Due to the fact that Marie matches both of her father’s cousins at 222M before Virginia’s crossover at 225M means that Marie has a maternal match in that area. Here is Marie’s Paternal Chromosome 2 filled in:

Marie appears to be Butlerless or Butler free in Chromosome 2. This is a good example showing that Marie got exactly half of her DNA from her father and half from her mother. However, when we consider her Paternal Chromosome 2, she does not get 1/4 from each of her paternal great grandparents. She got 0% from her Butler great grandparent. She also got roughly 1/3 from her paternal grandfather and 2/3 from her paternal grandmother.

Back to Uncle Naffy

This brings the story full circle. I started this Blog based on a few large matches with Uncle Naffy. Uncle Naffy’s family stories lead me to believe that he was related to the Crowleys. A Crowley married the first Irish immigrant Butler in my wife’s line. I identified the Paternal Grandparents in the visual phasing in this Blog based on the assumption that Uncle Naffy was indeed a Crowley descendant. So, ironically, Uncle Naffy’s match results lead to the mapping of my father in law and his sisters’ Butler DNA which lead to the conclusion that my Butler wife Marie had no Butler DNA in her Chromosome 2.

Summary and Observations

  • The maternal side of these Butlers chromosomes are easier to map than the paternal side due to lack of verified paternal matches
  • The paternal side match with Uncle Naffy has not been linked to a tree, so will have to be verified at some point.
  • Having cousin matches made it possible to fill in this map. Otherwise it would still have blanks.
  • While writing this, I may have found a Kerivan ancestor match. I will follow up on that and likely write another blog on what I find.
  • Once some good Kerivan and/or Butler matches are found, they will likely lead to other verified matches on those lines. It is difficult to break through and get those first identifying paternal matches.
  • I had thought that my wife’s DNA results were somewhat obsolete after getting her parents’ results. Now I see that I can map her great grandparents thanks to Blaine Bettinger’s instructive Blog.

Solving Joanna’s Mystery DNA Match with Visual Mapping

Recently I had a question from Joanna, who is part of a Frazer DNA Project that I’m working on. She has a large mystery match and would like to know which side of the family the match is on. Joanna is also interested in having her chromosomes mapped using Visual Phasing. Visual Phasing is a method that Kathy Johnston has pioneered using the DNA results of at least 3 siblings. Blaine Bettinger has also written a 5 part series on this subject.  Perhaps the mapping could help her find out what side of her family this mystery match is on.

Joanna’s Mystery DNA Connection with Mystery Vickey

Joanna’s siblings are Janet and Jonathan. I will check Gedmatch.com to see how the three siblings match up with Vickey.

vickeytojoanna

On Vickey’s One to Many list, I saw Joanna and Janet, but not Jonathan. You can see why Joanna is interested as her match with Vickey is 55 cM. I didn’t want to leave out Jonathan, so I ran a One to One between him and Vickey at Gedmatch;

vickeyjonathan

Jonathan does match Vickey, but he just fell off the bottom of Vickey’s One to Many List. The start of Jonathan’s match is at the start of Joanna’s orange bar above. His match with Vickey ends before his sister Janet’s match with Vickey starts. Now, in Joanna’s family we have a small, medium and large match with Mystery Vickey.

Visual Mapping of Joanna, Janet and Jonathan

As all the above matches are on Chromosome 13, it would make sense to start there. The first step is to compare the 3 siblings in the Gedmatch Chromosome Browser:

chromosomebrowserjjj

I then added crossover lines and attempted to assign the right sibling or siblings to the right crossover. This Chromosome was not simple. It looks like there are or could be close crossovers in three different places – around position 29, 33 and 98. In addition, something strange seems to be going on at the 72/73/74 location. That leaves only 2 crossovers which appear to be less than complicated. Those are: the first crossover which I have given to Jonathan and; the crossover at 62 which I gave to Janet.

Mapping the JJJ siblings

From 33 to about 73, Joanna and Jonathan have a Fully Identical Region (FIR). That means in that area, Joanna and Jonathan got their DNA from two of the same grandparents. One of those grandparents was on their Paternal side and one on the Maternal side.

chr13jjjfir

Above I’ve portrayed Joanna and Jonathan’s shared grandparents as blue and red. The next step is extending the blue and red bars. I’ll keep Jonathan’s two grandparents where they are as he has crossovers on either side. However, I’ll extend Joanna’s DNA from her 2 grandparents to the right. Should I extend them to 98 or to the end?

A single or double crossover at 98?

The simplest scenario at Position 98 would be a crossover assigned to Joanna. Joanna is the common factor in the first two comparisons, so that would make sense. In that scenario, neither Janet nor Jonathan would have crossovers at 98. However, that does not appear to go along with what we know. Recall above that I looked at Vickey’s match with Joanna and Janet:

vickeytojoanna

Why isn’t Janet matching Vickey for 55 cM as Joanna is? Something happened around Position 97. That something has to be a crossover for Janet. That sets a few things in motion. Now that we know Janet has a crossover, that means that Jonathan also has a crossover there. Our two options at 98 were either a single crossover for Joanna or a double crossover including Janet and Jonathan. My conclusion:

  • Janet has a crossover at 98
  • Jonathan also has a crossover at 98
  • Joanna does not have a crossover at 98

The immediate result is that I can send Joanna’s DNA over to the right side of the page:

joannamap

In the next step, I want to take advantage of the Joanna to Janet FIR and the places where siblings don’t match at all. This is where there is red in the Browser and no blue bar below. In these areas I will use two different colors.

firhir

The Irony of Vickey’s match

Next, I’d like to use Vickey’s match with Joanna and Janet set the crossover at position 98. We will now pick a ‘side’ to the colors. We will say that Vickey is a red match. In the area between 62 and 98 Joanna and Janet are in a Half Identical Region (HIR). That means that they have one grandparent in common our their four. The irony is that we are mapping this Chromosome to find out where Vickey fits in. Yet we are first using Vickey’s match to map the Chromosome.

vickeyonmap

Here I have added in Vickey’s matches with Joanna, Janet and Jonathan. Note that Janet’s match with Vickey starts right near her crossover. Before that, Janet matched another of her four grandparents (shown in yellow). Also Jonathan’s match with Vickey ends right by his crossover where he received his DNA from another grandparent (the same yellow). Now we have Joanna and her two siblings’ Chromosome 13 roughly mapped out. We have relative positions for their four grandparents. Can we now find out which of Joanna’s grandparents the Mystery Vickey is related to?

Finding Which of Joanna’s Grandparents Matches Vickey

In January, 2016, I wrote a blog about Joanna and her Chromosome 15. Here are the results I came up with (after I went back and corrected a mistake I found):

chr15frazermaprev

I show this to indicate the possible grandparents that Vickey could be related to. Chromosome 15 was also a much easier Chromosome to analyze. There were only three crossovers for Joanna and her siblings on Chromosome 15. The 90-97 area was where Joanna matched a Williams relative. From 67-92, Joanna and Janet matched Frazer relatives.

For Chromosome 13, however, the [maternal] Williams relative did not match with Joanna, Janet and Jonathan. So it will not be possible to make determinations on Joanna’s maternal side. There were, however matches on the Frazer side. Betty has not uploaded her results to Gedmatch, but she is on FTDNA which has a browser and gives Chromosome match locations.  Betty is our last chance to at least identify the paternal part of Joanna and her siblings’ Chromosome 13.  Fortunately, Betty matches Joanna twice and Jonathan once:

bettychr13

Now all I have to do is see where Betty’s DNA would fit on our Chromosome 13 Map.

chr13mapbetty

It looks like Betty has shown that Vickey is related somewhere along the line of Frazer – or more specifically Edward Frazer born in 1867. Here are a few notes on what I did:

  • I extended Jonathan’s Seymour DNA to the right as he had no Frazer match there where his two sisters did. They both matched Vickey. Then I added a blue segment above Jonathan’s yellow segment as Jonathan has a HIR with his two sisters at the end of Chromosome 13.
  • Question: Why didn’t I match Betty to the blue maternal side of the Chromosome? There is room for her DNA there also.
  • Answer: Imagine that I moved all the Betty matches up to the blue segments. That would leave a problem for Janet. In that scenario Joanna would have a blue match and Janet would also have to have the same blue match, but Janet didn’t match Betty from 75-96. Janet also did not match Vickey from 63-107. So that alternative scenario does not work out.
  • For more distant relationships, one would not want to make deductions based on lack of matches. However, with siblings there has to be an explanation as to why one sibling would have a match and the other would not based on the visual mapping.

Edward Frazer is wearing a top hat at the bottom right. In our Chromosome 13 Map, his DNA is shown as red and his Seymour wife as yellow.

frazer1867

The fact that Vickey matches on this Frazer line doesn’t mean that Vickey has to have Frazer ancestors. It just means that she and Edward Frazer must have a common ancestor. That common ancestor may be along the Palmer line, for example (Edward Frazer’s mother’s Line).

Betty to the rescue again

Let’s use Betty’s results to fill in the rest of Joanna’s family’s Chromosome 13:

chr13joannafinished

  • Note that Betty matches Joanna from 29-43 and Jonathan from 33-43. That tells me that Jonathan has a paternal crossover at 33. Because Joanna doesn’t have a paternal crossover at 33, that means she has a maternal crossover there. The rest I fill in using the FIR and HIR regions.
  • The smaller segments at the beginning of the Chromosome correspond with all the crossovers at the beginning of the Chromosome. There are 5 crossovers up to position 33. In Chromosome 15 that I mention earlier in the blog, there were 3 crossovers for the whole Chromosome. After position 33 on Chromosome 13, there are fewer, more spaced out crossovers which account for the larger segments of inherited grandparent DNA.

Summary and Conclusions

  • Visual Mapping can be fun and helpful in finding out where mystery matches come from
  • Without the help of Joanna’s 2nd cousin Betty, we would not have a complete map. We would also not be able to know which grandparent Vickey was related to.
  • If Betty’s results were only at AncestryDNA, we would not be able to do this analysis as AncestryDNA does not give detailed information on DNA matches. The fact that she tested at FTDNA helped us come to these conclusions, even though her results were not uploaded to Gedmatch.com
  • Joanna may know of more test results with known relatives that could help fill out the maternal side of Chromosome 13 or we may find out more in the future.