This month – another post on testing siblings and on using GEDmatch.
Sometimes, having siblings’ DNA tested is a real treat—such as when one of them connects to a distant cousin still residing in Ireland, just a few kilometers from where my ancestors came from. But sometimes—playing with sibling DNA can be a real challenge! Case in point: Chromosome mapping with three siblings. This is advanced stuff, yo.
To quote the ISOGG Wiki, “Chromosome mapping is a technique used in autosomal DNA testing which allows the testee to determine which segments of DNA came from which ancestor. In order to map DNA segments on specific chromosomes it is necessary to test a number of close family relatives. Ideally one should test both parents, one of their children, and a number of first to third cousins on both the maternal and paternal sides of the family.” (See the ISOGG Wiki page Chromosome Mapping.) Many of us, however, don’t have parents available for testing. If we have a group of three to five siblings who’ve taken an autosomal DNA test, there is another (albeit complex and time-consuming) way to identify which segments of our chromosomes came from which of our four grandparents. Genetic genealogist Kathy Johnston has developed and documented a methodology for this: scroll down the Wiki page linked above to see the section labeled ‘Chromosome mapping with three siblings’. This has been demonstrated graphically online by Jason Lee in “The Use of Crossover Lines to Determine Segment Matches with Grandparents Among Siblings—Visual Phasing” .
I used this technique to produce the chromosome 6 example shown in my February 2016 blog post.
Why do we want to map our chromosomes to grandparents? (Besides for the fun of it, and trying to figure out where that red hair or musical talent came from?) New people are taking DNA tests every day. When they turn up on our match lists, at least on a DNA website that gives us the chromosome and location of the match, we may be able to narrow down which ancestral line holds our common ancestor. Instead of comparing all 64 of my 4x-great grandparents to theirs, I may only need to use 16 of them.
A Cautionary Tale
[Spoiler alert – there’s a happy ending in the upcoming blog post: Chromosome Mapping Using Siblings – Part 2.]
So, I’ve worked on a few other chromosomes over time. Recently I ran into a hiccup that tempted me to title this post “A Cautionary Tale.” I’m not going to go into great detail on the steps involved; Kathy Johnston has explained (and illustrated) them far better than I could, and other bloggers may have demonstrated their progress already too. If you haven’t tried this yourself yet but are still interested, check out that Wiki page on Chromosome Mapping and associated links now. I’m going to just touch on the highlights of my conundrum here.
If you have a group of 3 to 5 siblings who have tested, you can choose a chromosome to start with—preferably one where you have confirmed DNA matches no closer than first cousins once removed or second cousins, i.e. where you know which one of your grandparent’s lines connects you to them. You’ll need at least one such cousin on your maternal side and one on your paternal side (more is better, as will become apparent shortly), and they need to match you or a sibling on the chosen chromosome.
Next, compare each of the siblings to each other using the one-to-one tool at GEDmatch.com, and copy the graphic depiction of that comparison so they line up one above the other, like this. Kathy Johnston recommends using PowerPoint.
We all know that we have two copies of each autosomal chromosome, one from Dad and one from Mom. We know that the DNA from each parent recombines before it is passed down — we may get a paternal copy of chromosome 11 that has a piece of Dad’s father’s DNA followed by some of his mom’s DNA. Our maternal chromosome 11 may contain some DNA from Mom’s mom, then her dad, then her mom again. (See the chromosome 6 example earlier if that helps illustrate the concept.)
The green sections in the image above show where two people match on both their maternal and paternal copies of chromosome 11. Yellow blocks show where they match on one but not the other (and don’t tell you which side is which.) The red stripe area, which doesn’t have the solid blue along the base, means that the two people do not match on that segment on either their paternal or maternal chromosomes. For example, maybe Brian got DNA from his two grandmothers at the beginning of that chromosome, whereas Margaret may have gotten DNA from her two grandfathers on that segment, so they don’t match each other there.
When two people switch from matching on one copy of chromosome 11 to matching on neither copy, or vice versa, it stands to reason that one of the people in that pair had a crossover event at that point. What’s a crossover? Dad has two copies of chromosome 11, one from Granddad and one from Nana. Those two copies are ‘recombined’ into one copy that has (usually) some Granddad DNA and some Nana DNA, which Dad then passes down to his children. (The same thing happens with Mom to create the copy that she passes down.) The location(s) in this copy of the chromosome where the DNA switches from Granddad’s to Nana’s, or vice versa, are the crossover points.
That’s what we’re trying to find. So, we take the above diagram and draw lines to try to determine where the crossover event occurred. GEDmatch will also give you the numeric start and end points from matching on one copy of the chromosome to matching on neither. It doesn’t, however, give the start and end points when going from matching on one copy to matching on both copies, or vice versa. David Pike’s utility “Search for Shared DNA Segments in Two Raw Data Files” can be used for that, if you want solid numbers and don’t want to just eyeball it.
Here’s what Chromosome 11 looks like with the lines drawn.
In addition to the lines, which indicate where a paternal or maternal chromosome crossed over from one grandparent’s DNA to another, a letter is shown, identifying which of the three siblings had that crossover event. For example, the farthest right line occurs at around position 130,000,000. (We’ll call it 130.) Margaret and Ann went from matching on neither copy of chromosome 11 to matching on one copy. Margaret and Brian also went from matching on neither copy of chromosome 11 to matching on one copy. Brian and Ann showed no change there: they matched on both copies before and after that position of chromosome 11. It seems likely then, that Margaret is the one who had a crossover event at this juncture.
If we look at the whole chromosome, it appears that Margaret had only one crossover event—that one near the end. That suggests that she got one copy from one of her parents with no crossover at all: it came from just one grandparent on her maternal or paternal side and got none from the other grandparent on that side. (This isn’t uncommon.) On the other chromosome, she got most of the DNA from one grandparent but a little, starting around position 130 from the other.
Plugging the numbers from GEDmatch into Kitty Cooper’s Chromosome Mapper yields this diagram for Margaret, limited to chromosome 11.
Now, the problem is this. Chromosome 11 was chosen because I know that Margaret has one-to-one matches with relatives connected through three of our four grandparents. For example, just on our maternal side, Margaret matches Grandpa’s relatives between 69 and 107. And she matches Grandma’s relatives between 109 and 116. So Margaret should have a crossover somewhere around 107-109. And there is a crossover there, but it was assigned to Ann, as it also appears to be where Brian and Ann switched from matching on one chromosome to matching on both.
If I just used these three siblings, eyeballed the results, and trusted that I had found and assigned the crossover points correctly, I’d be wrong. And if I didn’t have several cousins, close-but-not-too-close, already identified on this chromosome, I’d never even know I was wrong. And it might have led me to make incorrect assumptions about where new DNA matches might be related to me.
Fortunately, there is a happy ending. Stay tuned for Chromosome mapping with siblings – part 2.
Ann Raymont (c) 2016