GEDmatch Tricks and Tips

Question

Looking at the possible genetic matches to one of my ancestors, I initially found no match to any of them, but if I lower the number of cMs to 2 and the SNPS to 100, VOILA, we're related! I suggest anyone who is baffled by no connection when they think there should be one, lower those numbers on GEDmatch. This is especially true when (like in this case) the relation was born in 1745. The amount of shared DNA is miniscule, but it's still there.

Answer 1

There are really two issues here. 

One is that no match between two distant relatives does not prove they are unrelated genetically. The odds for finding a match between relatives drops rapidly by genetic distance. See the following link:

Cousin statistics (at ISOGG)

The second point is that lowering the matching criteria to 2 cM really proves nothing either way, even if a "match" is found. The links below explain this in great detail. 

When is a DNA segment match a real match? IBD or IBS or IBC? (by Kitty Cooper)

Identical by descent (at ISOGG)

To summarize, you won't share a significant amount of DNA with every relative and you shouldn't expect that. Lowering the DNA matching threshold also dramatically lowers the confidence that your shared DNA actually means anything.

Comments

Thanks Bill, for the detail and links. I' not relying on any of the information from the altered parameters, but learning more is always a good thing. A good example of how we carry genetic material from our parents and theirs before them is a recent relation who found me. She sent me a message through 23andMe because when her results came back, I was her second closest match. The first was a cousin on her maternal side whom she already knew.

What she didn't know was who her father was. Her mother had passed years before and she didn't even have a first name. So through a lot of searching on my Ancestry tree, which is far more developed than what I've accomplished here yet, we were able to narrow it down to my paternal line. She did another test through Ancestry so we had more people available to triangulate. In the end, we found her father. The distance in our connection is what I find so fascinating. As it turns out, her great-great-grandfather was the brother of my grandfather. The ages are part of what confused us. I'm 19 years older than her so we thought her father's generation would be no more than one generation greater than mine, but obviously, that wasn't the case. Still, we share 2.04% of our DNA and 152cM across 6 segments.

Answer 2

I got all of my DNA from my parents, so any DNA I share with a match should have come from one of my parents. That parent should be a match too. Yet I have found many segments of length 7-10 cM that I share with a match, yet neither of my parents does. The explanation is that by chance, some combination of my parents’ DNA is identical to my match’s, but we didn’t get that DNA from a common ancestor.

Segments shorter than 5cM are estimated to be “by chance” like this more often than they are real.  Some of them will be real, but there is no easy way to tell which are true.  Better not to lower the GEDmatch thresholds since otherwise you may be led down a very long path to nowhere. For instance, if you find a third person who shares one of the very short segments you mentioned, I would absolutely not assume that  person has John McQuown or even one of his ancestors in his/her tree.

Sharing no significant DNA segments doesn’t mean you’re not related.  It happens even for second cousins once removed, every so often. When it happens, take it in stride and work on the matches with whom you do share a bunch of DNA.

Comments

Yes, I think it's important for folks to realize this not only for the reason you describe but also so that they don't immediately think of it as a red flag or let their mind go to a non-paternity event.

---

Good points to be kept in mind, but when you say you got all your DNA from your parents, it's not altogether accurate. While technically that is where it comes from, there is a cascade of DNA that is passed down through multiple relations, and we get different amounts, and sometimes none, from each of them. Still, going down a garden path of incorrect information is not the desired outcome!

---

There was nothing inaccurate about my statement.  You offer a different, equally accurate perspective.  But for how I used the statement in my comment -- indicating how you can find verifiably false matching segments -- the fact that DNA also came from grandparents and more distant ancestors is not relevant.