Segment Comparisons, Perspective and Inheritance

I wanted to post some examples of the differences between close relatives as they relate to more distant relatives and touch on inheritance and the luck of the draw. The reason for showing these is to give some sort of visual display of the ways in which related people match each other in different ways.

My Dad and I

I'll start with the one that I've had the most comments on at 23 and me. When people compare me to my dad they see something like this:

given the small format, you may want to click these to get a closer view. Above is my comparison to my dad. The comments I normally get from people are along the lines of "how can it be that you are exactly like your dad". 

The key is that my dad gave me half of his DNA. He built half of every chromosome I have (except the X). So I'm half identical on every chromosome. What it's easy to forget is that most chromosome browsers show you "half identical" segments...not fully identical. 

When you have a match to another person (unless they are a full sibling) you will normally only have half identical segments in common.

What I'm showing here is that for me, there is the opportunity to have a match to my paternal family on every chromosome, in every position. Even though every chromosome is covered, it's only half my DNA.

My Dad, his Maternal Half Sister and their Maternal Uncle

Okay. Time to add some complexity. Here is the broad overview of my dad, his uncle "R" and his half sister "C". This is from my dad's perspective his half sister only has DNA from their shared mother's family. His maternal uncle is his mother's brother so he has the same family in common. 

I'm going to close in on some of the chromosomes at the top to illustrate some differences. I apologize if you have trouble telling blue from green. I didn't have any choices for colors at 23 and me. My dad's uncle R is in green. Half sister C is in blue.

Things to note when looking at this. Notice that even though R, C and my dad all share the same very close family and all the same ancestors on my dad's maternal side, there are differences in how R and C match my dad. These matches represent DNA coming to my dad from his mother. 

You can see that there are places, like on chromosome 1 where sister C matches my dad but Uncle R does not. That makes sense because uncle R and my grandmother were not twins, they got different DNA from their two parents. 

You can see that there are places where uncle R matches but sister C does not like on chromosome 5 and a good chunk of chromosome 4.  Again here uncle R and my grandmother have the same DNA and my grandmother handed some segments down to my dad, but not his sister. This also makes sense because my dad and his sister are not twins. 

They both got half their mother's DNA, but the halves they got were not identical. The halves were randomly generated from all the DNA available to my grandmother.

Also notice these pictures compared to my match with my dad. Even my dad's uncle combined with his sister is not enough to recreate the DNA my grandmother gave my dad! There are maternal relatives in those white gaps that don't match his sister or his uncle.

From the Sister's Perspective

Now I want to look at these same matches from my dad's sister's perspective. The same three relatives all share the same family. The matches are similar in many ways which makes sense, but they are different in striking ways too. My dad here is in Blue and R again is in green.

Compare and contrast with the close up above. Sister C has a remarkably bigger segment in common with uncle R on chromosome 3. Where my dad has an amazing amount of white space, she is a half identical match to uncle R.

What is important there is what you don't see in my dad's close up. Remember that his mother fills up the entire chromosome for both my dad and C. It's not that he doesn't have relatives of his mother there, it's that they will most likely be different relatives than C and R have in that same spot.

For example. C and R may carry a large segment of DNA from one maternal relative while my dad carries a large segment there from another maternal relative. My grandmother had both to give, my dad got one and my aunt got the other. In really simplistic terms, maybe my dad matches his maternal great grandmother there, while C and R carry DNA from the great grandfather...or something to that effect.

What's the Point of all this?

The point is that even closely related people who share the exact same set of close ancestors and get huge amounts of DNA from those ancestors, do not get the exact same DNA from those ancestors. Uncle R is not definitive for my dad's maternal matches. Neither is sister C. Together they come very close, but still not close enough to cover all the ground. In those cracks are the exact same ancestors shared by everyone in the group, just broken up in a different way.

Somewhere, another cousin may fill the gaps in. Perhaps one of my dad's first cousins through a different aunt or uncle will hold the keys to those white areas. They will have matches to the same families but in different spots on different chromosomes.

Here is my last example image, I promise. This is from the perspective of sister C again. My dad is blue, Uncle R is green and common cousin P is purple. 

P is related to C and my dad's great grandmother Clara Beadle. Clara is uncle R's grandmother. P shares about 206 total cM with sister C, but I'm going to focus on chromosome 22 for a minute because it's a perfect illustration. 

Uncle R in green overlaps my dad a little there in blue. Both do not overlap with P, all are related to Clara Beadle in some way but that doesn't mean they will always match each other where P does because they all got different DNA from the Beadle family. Each of these matches could represent Beadle DNA. P cannot be the definitive check for Beadle DNA because he doesn't cover enough ground. 

Here is what I mean by P can't be definitive for Beadles in my family:

P shares 206cM on 8 segments with sister C.

P shares 156cM on 6 segments with my dad.

P shares 145cM on 10 segments with uncle R.

The Beadle family is about 1/4th of uncle R's DNA and yet P shares the lowest amount of DNA in total with R! My grandmother has more in common with P and handed more of those common segments to her children C and my dad. Uncle R is not less related to his Beadle grandmother than his sister was, he just got different DNA from her!

Updated Y Results and a Psychological Shift

Although you have to question your mental health when reviewing how much time and effort you put into genealogy, I have not been more aware of my own perception change in a long time. Today, the results for my Elmer Y DNA recruit down the Ed2 line were revised. I wrote about his results and my interpretation of them last time.

Today, while comparing the results for another Elmer tester, I noticed a difference in the genetic distance from my recruit to me. Where there was a GD of 2, now there was 1.

His odd mutation at GATA-H4 had been wiped away and the more standard value of 10 put in place of his 9.

So now he shared the non-modal value of 10 there like the other Elmers and I, but he still carries my mutation of 29 at DYS449.

All of that means that I have my closest ever Y match. My interpretation of his results discounted his GATA-H4 value, because it was so different from the others. So really, nothing has changed about the match in my eyes. It's still my closest match because of that shared mutation at DYS449. I feel it is the most likely candidate for a paternal line for my branch of the Thompson family. It's still my best bridge between the Elmore family in Illinois and Edward Elmer in connecticut.

Psychologically though, over the course of the day, my perception has changed about the match quite a bit. In my own mind it went from the most likely match to an absolute match. Dangerous ground for a person who is most often wrong, but still there it is.

Now this minor change has me thinking the autosomal match to the Elmore family is better than I thought. More solid. Like the one paternal segment and unknown second segment in that ATDNA match are, all of the sudden, an excellent sign post rather than a vague possibility. Even though it really says nothing about those results at all, it has me feeling degrees more hopeful that I've done this bit right and my gamble has paid off.

Coming down from my high, this Y match is not definitive for the autosomal match. I would need a Y from a closer Elmore relative to triangulate with or at least one more autosomal test from a related Elmore to triangulate that.

Still though, optimistically, I feel certain that this Y result means I am related to Hezekiah Elmer 1686, son of Edward Elmer and Rebecca Fitch. Certainty is not something I'm used to.

A New Year and Some Encouraging Results

The Y results are in for our last Elmer Y recruit. These are the results I was waiting for in my last post about Elmer and Elmore autosomal results.

The autosomal results for this firs contact were "no match". Not matching is hard to gauge because it's possible for genetic relatives to not match each other at all. When you don't match a close relative that can be a red flag, but not matching someone who shares a relative in the 1700s is more likley.

I described the layers of expectation for the Y results like this:

In layers of expectations, I would expect them to be a good match to the other tested Elmer branches. I would hope that they might share one of my variations from the Elmer "norm".

You better believe that I was holding my breath looking at that Y match. He was definitely a good match to the group but he was still two markers off from me! At that point, my heart kind of sank. All the Elmers who have tested beyond the first twelve are a minimum of two away from me and it's the same two every time.

I carry a one off mutation at DYS449 and DYS567. FTDNA lists those two places as likely mutations at 37 STRs for people who are related, but their stipulation is that they also share a surname. I do not share a surname.

I was stunned though to find that the Elmer I suspect is on my Y line shares my mutation at DYS449. He mismatches me in two spots and he mismatches the other Elmers in two spots (one of them being the spot that matches me).

Time for some visuals. First up is the Y 12. I've got the men organized by their apparent or documented branch of the Elmer family through Edward Elmer's sons John, Edward2 and Samuel. I put myself with L (gray bar near bottom) in the Ed2 category. Fairly similar at 12 we have most of the telltale cumberland markers from the cumberland B modal (green bar at bottom). The modal for the larger group in Z18 is the green bar above.

Two of the documented John lines have some interesting mutations on normally slow moving values.

In the second set, everyone has the telltale cumberland B 458.2. What grabbed me is that where I expected a mismatch on DYS449 like all the other Elmers and the modal for cumberland and the modal for Z18, I now have a match! Both L and I are 29 at DYS449.  So L matches me where the others don't. What is odd here is that L has a mismatch with me because he's broken away from all the Elmers on GATA-H4 which is usually a more stable STR. So where he mismatches me here..he mismatches everyone.

Last set. Again an odd break away for the "John" line Elmers on some slow moving values and they have also moved away from cumberland and Z18. The big orange 18 is me. Here I mismatch all the Elmers still along with the Cumberland and Z18 modals. The final DYS463 value is included to show those that have tested for it. We have one John, one Samuel and if I can be counted one Edward2 showing a 23 there against type for both cumberland and Z18.

One thing I want to note is that the John group seems to be wild. They mutate away from each other. They have standard Elmer markers in common but at 37 they don't seem to have some "John" pattern.

Meanwhile the Samuel group is super super stable. They form the core group to compare others to. They have the most STRs in common with everyone. 

The Edward2 group, well...

I've spent a lot of years with my results, trying to gain value from what I have. I feel like I've earned some good returns. This is not perfect, but if my suspicions are correct, L and I meet up at Hezekiah Elmer born 1686 the son of Edward Elmer and Rebecca Fitch and grandson of Edward Elmer. 

There is a lot of time there to gather mutations and the "John" group shows how much variation there can be. I am most interested though in where L and I are similar. Not just wildly mutating by ourselves, we have a mutation from the "norm" in common.

L could have randomly gotten the 29 at DYS449, but I'm going to take it anyway. I didn't randomly recruit L. I picked him because I suspected he and I were on the same line and I was hoping he would share one of my variants from the other Elmers... and he does. That would mean that Hezekiah or Edward2 is the originator of our 29 there. My Elmer mismatch at DYS576 and L's really odd mismatch at GATA-H4 would then be marks of our individual lines at some point after that.

I was concerned about trying to match up with someone who was that far back. There isn't a lot of time between Edward1 and Hezekiah. Mutations can happen anywhere at any time, but still only a couple of generations from the source and I'm hoping for a common STR that is not shared by the majority of Elmers! The thing is...I got it. I actually got it.

He has one of my mutations but not the other, he's part way there, just like someone who shared a very old common ancestor with me.

One way to feel better about it would be to get another person related to Hezekiah tested to see if they also carry the 29. What if they did not though? Mutation is not a one way street. G in the "John" Elmers has mutated back to the old Z18 modal twice! A member of the Hezekiah branch might drop the 29 in favor of some other value.

Another way would be upgrading both L and I to 67 and 111 to see if we maintain the closeness.

Yet a third way, and the one I might save up my pennies for, is to run both of us through Big Y to see which SNPs we share and how they layer with other Elmers who have done it.

Leaving off on a hopeful note. I have a good autosomal result from my targeted Elmore family and a nice Y match on their line back to Edward Elmer. I may even have identified one STR that we could assign to an ancestor.  

That is not too shabby.