In my previous Blog, I created a very useful spreadsheet thanks to Steven Fox. This spreadsheet has my five siblings plus me with comparisons to each other, cousins and all I need to visually phase six siblings. In that Blog, I started by mapping out the X Chromosome.
On to Chromosome 1
Chromosome 1 could be tough. This Chromosome has the potential for the most crossovers. I tried showing some crossovers, but it was a bit of a muddle:
I have worked on phasing my DNA using raw DNA and have had help from M MacNeill. MacNeill came up with this for three of the siblings in my family:
This shows a maternal and paternal crossover for me in close proximity at about 23M. MacNeill showed my maternal crossover at 23.1M and my paternal crossover at 23.3M at the left side of the top row. That is fairly close.
Here is a problem area on the right hand side of my sibling comparisons:
In the D&S comparison, see how quickly the comparison goes from FIR to no match. This is likely indicative of my maternal and paternal crossover. Above, the F&L comparison, there is a slightly more gradual change from no match to HIR to FIR. I don’t know if that indicates a maternal/paternal crossover for Lori or if my double change is giving Lori a double change.
Then, on the right side of the Chromosome, I see this:
At around position 204M, F&H goes from no match to a FIR fairly quickly and D&L goes from HIR to no match very quickly. It looks like I could have another Lori (L) crossover in that area.
Here I tried some visual phasing, but, it is not right:
Here is how the segment map compares:
The comparison is not bad in the middle of the Chromosome where I started the visual phasing, but it goes south quickly on the right. The V&D comparison on the left side of the segment map is off also.
Using Cousin Matches
My though at this point was to use cousin matches to get this map into better shape.
At the bottom right, Lori’s last match with Joyce ends at 203M, Heidi’s ends at 205M and Sharon and Joel continue on. Sharon’s Match with Joyce starts at 190M and my match starts at 182M. That tells me that Sharon has a crossover at 190M, Lori at 203M and Heidi at 205M. This is just on the paternal side as Joyce is a paternal match.
A Cousin Match Spreadsheet
I like the idea of using cousin matches, so I copied this spreadsheet and made another called Cousin. The good thing about putting in matches based on cousins is that when you can see the crossovers, you can tell which side they are on. Then on the other side of the crossover, the DNA will come from the spouse of the match. For example, above when my siblings match Joyce, that is on the Hartley side. So to the left of Sharon’s Hartley match with Joyce, the DNA would be from Frazer (my paternal grandmother).
Here is how I mapped out my siblings’ Hartley match with Joyce. On the other side of the paternal crossovers, I can add in Frazer:
After comparing the paternal matches with the gedmatch comparisons, I can make some more deductions:
For example Sharon has no match with Lori in the middle segment. that means that Sharon must have a Frazer segment there where Lori has the orange Hartley segment.
Adding Maternal Cousins
My cousin Catherine represents my Rathfelder maternal grandfather. I’ll add in some Rathfelder from Catherine:
Where Jon does not match Catherine or Rathfelder, he must match Rathfelder’s wife who was Lentz. This is true for Joel also in the second set of matches. This change for Joel (D) must happen where I have his crossover.
This gives my family four grandparent segments by just two cousin matches.
Two Missing Paternal Crossovers for Lori and Heidi
Above, I had mentioned that Lori had a paternal crossover at 203M and Heidi at 205M, but I forgot to add them in. Here are the ones I forgot along with the ones I didn’t:
This shows how crowded things can get with a six sibling comparison. This is what these crossovers result in on the map:
That configuration results in this Segment Map on the Fox Spreadsheet on the right versus the original gedmatch comparison:
Above, I am just comparing the segment that goes from L to L. All the differences are on the right hand side and they all have to do with Heidi. I suppose that could mean that I have too many crossovers for Heidi, not enough, or have deduced the wrong thing from the crossover I have. My guess is that Heidi has an additional maternal crossover.
Based on the Gedmatch comparisons, Jon (F) and Heidi should match on both grandparents. If Heidi had a Lentz in her last column, then she would match Jon.
Now the segment map agrees with Gedmatch:
The problem was the 5 crossovers in a row. I find it interesting how a comparison with the segment map and Gedmatch pointed right to Heidi as the place where the problem was.
A Lentz Cousin Should Fill Out the Right Hand Side of Chromosome 1
Here is how Judith matches my family:
I see three maternal crossovers on the right. One is with Jim (V), then Heidi, then Lori in the last row.
When I fill out the missing segments, I get this:
Using the Fact-Checker Segment Map
The good news is that I have the area of discrepancy narrowed down to one person again:
I did something wrong with Jim (V). Jim is supposed to match Sharon in the Last segment. That means that Jim needs another paternal crossover from Frazer to Hartley. All I had to do was compare Jim to Frazer cousin Paul to find Jim’s crossover:
Actually I had to compare Paul to Jim and Jon to find Jim’s crossover. Here is my expanded list of crossovers:
Here is the extra Jim (V) crossover added before Lori’s last crossover:
There would have been a lot more trial and error without the cousin matches that I have. Now Gedmatch and the segment map appear to match:
Also, I am liking Steve Fox’s Segment Map.
The Most Difficult Chromosome
Chromosome 1 isn’t necessarily the most difficult Chromosme to visually phase but it has the most potential to be the most difficult one. Looks like this Blog will just be on Chromosome 1.
Here’s a Hartley cousin on the left side of the Chromosome:
This will take care of one of my crossovers and one for Lori. It looks like Lori’s crossover should be ahead of mine:
I also carried some of Jon’s segments over as he shows no crossovers. It looks like I need another crossover for Lori:
Jon (F) has no match with Lori in the first row comparison. Then he has a short HIR and then FIR. My assumption is that Jon has no crossovers on either side of his Chromosome 1. That means that Lori, must have two crossovers.
Final Fact Check with the Segment Map
Something is off in the V Section which is Jim’s. My solution so far, has been to add an extra crossover:
Now I have everything except a tiny little area for Jim:
If only I had a cousin match in that area. It looks like the first V is at 35M and the second is at 41M. Jim’s match with Joyce should work.
Joyce matches Jim (V) up to 38M which is past the area of his first crossover at 35M:
The Finished Product
Using the merge function beautifies the map:
Wait maybe not. Here is a global fact check using the Segment Map:
How did I miss this before? Again my fact-checking exercise has narrowed down the problem to one sibling. This time it is Sharon. I have also narrowed the area in which the problem is: between S and V.
I see the problem:
Sharon is third from the bottom. I have her bottom maternal crossover from Rathfelder (purple) to Lentz (green) at V instead of at S. This should fix things:
That should do it:
This shows that there are no gaps where any of the grandparents’ DNA is missing between all six siblings.
Summary and Conclusions
- First I made some rough crossover lines and assigned them to four siblings
- Second I went to the cousin match map and mapped a few cousins onto the spreadsheet
- Using these cousin matches I could see where some of the crossovers were. The advantage of this is that I already knew which side (paternal or maternal) the crossovers were on. On the other side of the crossover, I knew that the person got their DNA from the corresponding spouse.
- I extended these matches using the gedmatch sibling comparisons
- I checked my results using the segment map and made corrections
- I found where there was a quick change from no match to FIR or FIR to no match in a sibling comparison, that it often meant a person had a paternal and maternal crossover in close proximity.
- The use of cousin matches greatly sped up the process of visual phasing.
- By using cousin matches I skipped using the G1-G4 codes and went directly to actual grandparents. This method seems more intuitive to me. However, it only works with good cousin matches.
- This also went quickly because I had done it before. Because of this, I already knew I had a double crossover near the beginning of the Chromosome. I also believed that my brother Jon had a full pair of DNA from his paternal grandmother and maternal grandfather. However, I basically recreated the map from the beginning.
- It is satisfying to see the map for all six siblings and how they got their DNA from each of their four grandparents.