Would we all have the same common ancestor?

[FionaCal]

Hi all,

I've noticed on my list of matches that there are around 30 people, most of them in the first couple of pages out of my 600 matches, who not only all share the same segment of the same chromosome as me but they also all appear "in common with" each other.

Does this mean that all of us from that group share the same common ancestor?

Then what does it mean if some of the people in that group are an x-match and others aren't?

Intrigued!

Thank you!
Fiona

[Germanica]

Hi all,

I've noticed on my list of matches that there are around 30 people, most of them in the first couple of pages out of my 600 matches, who not only all share the same segment of the same chromosome as me but they also all appear "in common with" each other.

Does this mean that all of us from that group share the same common ancestor?

Probably, but be careful the segment you all share doesn't cross the centromere. My understanding is shared segments that cross the centromere, even if they are long enough to normally register as IBD, have a higher chance of being IBS.

Then what does it mean if some of the people in that group are an x-match and others aren't?

Just that some inherited X-DNA and some didn't. Maybe for some, it passed through more than one male in a row, meaning X-DNA couldn't have been inherited, but the other autosomal DNA was.

[FionaCal]

My understanding is shared segments that cross the centromere, even if they are long enough to normally register as IBD, have a higher chance of being IBS.

Thanks for your reply. Sorry, what are IBS and IBD?

[keigh]

Identical by descent (DNA that shares an ancestor) or identical by state (the DNA just looks similar with no ancestor involved) is how I've had IBD and IBS explained to me.

[John McCoy]

What is the source of the opinion expressed earlier in this thread that matching segments that cross a centromere are more likely to be spurious? I don't see the logic in that statement.

[mabrams]

There is a Matrix command that you can use, and verify that the people are actually related to each other. You can only try 10 people at a time, but you can see if the idea that everyone is related is a reasonable idea. It doesnt verify that they are all related on the same chromosome region but it would seem plausible.

Another concern is that these people are what is called by some, such as AncestryDNA, a pile-up. FTDNA does not check for pile-ups. AncestryDNA removes these segments, some of which are probably valid, and then often removes the match designation.

A lot (most) of X matches are very tiny and you shouldnt pay much attention to them. I have matches with tiny Xs to my mother and larger autosomal matches to my father.

[Germanica]

What is the source of the opinion expressed earlier in this thread that matching segments that cross a centromere are more likely to be spurious? I don't see the logic in that statement.

I believe it's from Genome Mate, I recall reading that is why Genome Mate has an orange alert whenever a shared segment crosses the centromere, to alert you to the fact that you should be cautious regarding it.

This isn't from the creator but:

http://genealogypuzzlesdna.blogspot....ation-and.html

"When a segment crosses a centromere, there is a greater chance the segment is a mismatch or IBS (identical by sequence) rather than IBD (identical by descent). When this occurs, the ICW group on the graph will be an orange color to alert to the possibility."

[John McCoy]

Interesting! But still not citing any evidence! I have a feeling there is a hole in the argument someplace. Yes, there is less recombination (per millions of base pairs) around the centromere, and fewer SNP's to work with, but the odds of multiple crossovers occurring within such a region seem way too small to present a significant confounding factor for the matching algorithms.

[Germanica]

Interesting! But still not citing any evidence!

If advanced genetics software like Genome Mate isn't evidence, I don't know what is.

Download the User Guide for Genome Mate: http://getgmp.com/Download/UserGuide.zip

Found on page 127, it says:

"Note: Genome Mate Pro considers a segment that crosses the centromere to be a “Potential False Positive”. This is noted on the Chromosome page with an orange question mark in the “Side” column."

I have a feeling there is a hole in the argument someplace. Yes, there is less recombination (per millions of base pairs) around the centromere, and fewer SNP's to work with, but the odds of multiple crossovers occurring within such a region seem way too small to present a significant confounding factor for the matching algorithms.

Perhaps multiple matches does make it more legit, but I haven't seen any comments about that anywhere. I am just relaying what I've read. Note it says "potential false positive", not definite.

I'm not seeing any evidence from you to the contrary?

[Germanica]

Here's more evidence, from FTDNA itself:

https://www.familytreedna.com/learn/...nt-centromere/

"How does the Family Finder program account for the centromere?

Our bioinformatics team treats centromeres as SNP Poor regions. We use a hard stop for matching across each centromere. This assures that DNA segments (blocks) are truly Identical by Descent (IBD) and not Identical by State (IBS). If you seem to have an IBD segment that crosses the centromere, we evaluate it as two segments, one on each side of the centromere."

[John McCoy]

You are quoting someone's method, but not evidence that there is actually something happening that needs to be corrected. I'm asking what is the evidence that matches involving SNP-poor but also recombination-poor regions are inherently too risky to treat as real.

[georgian1950]

You are quoting someone's method, but not evidence that there is actually something happening that needs to be corrected. I'm asking what is the evidence that matches involving SNP-poor but also recombination-poor regions are inherently too risky to treat as real.

Good point. I find a lot of bad assumptions in genetic genealogy which are backed only by previous bad assumptions.

Jack Wyatt

[Germanica]

You are quoting someone's method, but not evidence that there is actually something happening that needs to be corrected. I'm asking what is the evidence that matches involving SNP-poor but also recombination-poor regions are inherently too risky to treat as real.

I never said that it is inherently too risky to treat as real. I just said it COULD be and I think I've provided ample evidence as to why. Three different sources, one from FTDNA itself, confirming what I said but you just keep ignoring them like I didn't post them, and now you're ignoring my request for you to post evidence to the contrary. Not my problem if you can't accept what is staring you in the face.

[DavidFB]

Does this mean that all of us from that group share the same common ancestor?

This is the premise of ancestry research - that the paternal (Y) and maternal (mt) lines have distinct mutations that can be traced back to population groups in specific regions.

In other words, a common ancestor at a series of specific times and places going back about 100,000 years. The further back you go, the more common/ the bigger the included group.

The deep ancestry you see on services like National Geographics Genographic project show a map of the migration routes.

Notably, they're not suggesting an Adam and Eve, only that out of a given population, one persons DNA is what was carried forward.

Also note that this is true of both the Y and mt lines so not a single ancestor but a series in both lines.