Crumley Kinfolk

Just for fun, let’s look at some Crumley folks who are related.  When you work in the genetic genealogy field, people are forever sending photos of someone and saying “doesn’t this person look like that person?  Do you think they are related?”  Or, the most common, “this is my great grandmother – do you think she looks Native American.”

I am forever telling people that phenotypical resemblances are really not good indicators of relatedness, but it’s so difficult to believe when you’re looking for that needle in the haystack and it’s the only tidbit you have.

I did it myself when I found Lee Devine and discovered that not only was he deceased, but he had no children, so my chances of ever finding out definitively if he was my half brother are forever gone.  I reverted to picture comparisons, because it’s the only tool I had at my disposal.

So, let’s have some fun with this.

Take a look at this photo.  These men are unquestionably related.  The question is, how closely?

crumley kin

So, how do you think they are related?

If you said brothers, you’re in good company.  They look like brothers, but they aren’t.

If you said uncle and nephews, you too would be in good company, but nada.

Cousins maybe?

Well, yes.

These are all Crumley men, left to right, John, Ken, Jerry… and Donna, the daughter of one the men whose job it was to keep them in line that day.  Fortunately for Donna, she doesn’t have the signature family beard!

Years ago, the Y DNA tests through the Crumley DNA project confirmed that these men share a common Crumley ancestor, but despite appearances, they are much more distantly related than you might think.

Not first cousins.

Not second.

Not third.

Not fourth.

Not kidding!

The common ancestor of these men is John Crumley, born about 1737.

Yes, I know how much alike they look, but looks can be deceiving – or encouraging – and looks are not an accurate predictor of relatedness.

John and Ken are 4th cousins once removed.

Ken and Jerry are 5th cousins.

John and Jerry are 5th cousins once removed.

Their pedigree chart is shown below.

Crumley kin pedigree

Not quite what you would expect by looking at the picture.  As someone once said to me, “If you look at a picture long enough and hard enough, you can see anything that you want to see.”  Touche!

The Crumley DNA project at Family Tree DNA has embraced autosomal DNA testing, so all three of these gentlemen have taken the Family Finder test.  Knowing that their Y DNA matches (with a mutation or two), and having identified their common ancestor, let’s see if their autosomal DNA matches as well.

At Family Tree DNA, one must meet a 20 cM total DNA matching threshold, and an individual matching segment threshold of 7cM in order to be listed as a match.  Here’s how they matched, or didn’t.

Jerry John Ken
Jerry Self Yes No
John Yes Self No
Ken No No Self

Needless to say, if we didn’t already have the Crumley Y DNA results, this might have given Ken a bit of heartburn – but no need.  It’s not uncommon for distant cousins to not be shown as matches.

Fortunately, all three gentlemen also downloaded their results to GedMatch, where we can adjust the matching threshold.  In some cases, the 20cM total precludes a match, and in some cases, the 7cM segment precludes a match, so let’s see if these gentlemen match at GedMatch using a lower threshold.

At GedMatch, I ran all 3 gentlemen against each other using the threshold of 300 SNPs and 3 cM and then put their results into a common spreadsheet.  I also deleted the duplicate entries, because for every Ken to John match, there is also an identical John to Ken match.

You can see on the spreadsheet below that John and Jerry match each other, just as Family Tree DNA said.  They share not one, but two large matching segments of over 16 cM.  Not bad for 5th cousins once removed.

Crumley kin gedmatch

You can also see that Ken matches both Jerry and John, but not on any segment over 4.9 cM, which precludes matching at Family Tree DNA.  However, Ken exceeded the 20 cM total match threshold with both Jerry, at 51 cM and John at 35.8 cM – but a match has to exceed both thresholds to be counted as such.

Especially within known family groupings, a non-match doesn’t necessarily mean the individuals don’t share any DNA, it may just mean that there isn’t enough cumulatively (>20 cM) or the segments are too small to put them over the threshold (7 cM).  That’s the great thing about GedMatch, you can adjust your own thresholds.

Are all of these segments valid, meaning are they identical by descent?  Most likely not.  Are some valid?  Very probably, especially given that we know that these men unquestionably do share a common ancestor – thanks to their Y DNA.  Could we find out more?  Yes, we can, if we have more cousins to compare against.

And, as luck would have it, we do, another 40 or so….but that story will have to wait until the Crumley DNA Study is ready for publication!

Thanks to Ken, Larry and John, my Crumley kin, for DNA testing and allowing us to tell their story and share their picture.  You can see by the smiles on their faces that they are truly enjoying their kinship – and that is really what matters.  Genealogy and genetic genealogy has the ability to reunite families separated by more than 200 years and 6 or 7 generations – and that’s exactly what has happened with our Crumley kin.

18 thoughts on “Crumley Kinfolk

  1. Roberta, thanks for your insightful articles on Genetic Genealogy. While having tested early, I have only recently begun to work in the area. I have learned so much from you. Thanks for sharing your knowledge.

    Chuck Haine

  2. Roberta,
    My name is Farrell Crumley. I’ve recently started looking into my geneology, first by creating a family tree on Ancestry, and then, just this month, getting my DNA tested through Ancestry. I believe I’m descended from James Crumley. I didn’t know about the Crumley DNA project until just today. You mention there is a way to transfer my Ancestry DNA results to your project. I would like to do that & find out more. Please email me the info. (I would also be willing to get tested through your preferred company if that would be better. I’ll await your response.
    Farrell Crumley

  3. I have a brother, also a 6th cousin 2x removed, and an Englishman whom we suspect is an 8th cousin to me and my brother. We have all taken the AncestryDNA autosomal test and I have uploaded them to GEDMatch. I find that we four match only at 4-5 cm ~300 SNP. on one Chr 15 segment. We may have another 1st cousin, and 7th cousin joining in soon. i’m new at this but I hope to do one-to-many at GEDMatch for each of us at 4 cM 300 SNP thresholds, import the four search results into Excel, and then do a logical AND between everyone to see who remains on the triangulation. If only 10 or so out of the million GED kits triangulate on this segment, then I naively would say the chances of an IBS (is that the common ancestor by chance?) would be negligibly small. Your suggestions? Thank you. With 8th cousins, the 7 cm thresholds are too high I suspect to be used.

    • I don’t generally like to use segments smaller than 500 SNPs because they are more likely to be IBC. That isn’t to say you should discount this. Just be aware and look for corroborating evidence.

      • Thank you very much. I’ve just started this study after making contact with the supposed 8-g-gson of my 8-g-gfather. His pedigree is well established/documented back to his 8th g-gfather (who was a member of the English peerage). I have found a single matching segment among four of us (including 3 of my US cousins – one a 6th cousin 2x removed) I’m attempting to establish if we four share the same 8th g-gfather. My 6th g-gfather immigrated to the colonies in ~1740’s, and oral tradition states he was the g-son of this nobleman.

        Of course, after 8 generations the matching segment we four share is small (both cM and SNP). I believe my only recourse is to do GEDMatch one-to-many for we four, save the (GEDMatch one-to-many can produce 100,000 I’m told) results to a spreadsheet and then due the filtering (probably with LabView software for convenience). By filtering, I mean select only match on the same segment. The probability of a match by chance should then be proportional to the resulting matching kits, divided by the total number of GEDMatch kits in their database (now 800,000) So, if after filtering on this single segment, there only results say ~10 kits for which we all match this segment, the probability of it being a random match would be ~10/800,000. We shall see! 🙂 A few other documented but distant cousins may be recruited as well.

      • Thomas, I am a Crumley relative that traces his roots back to a James Crumley born around 1711 or 1712 either in England or Ireland. He appeared in Pennsylvania in the 1720’s & then in Virginia in the 1740’s. While I do not deeply understand Dna testing, if I am a distant cousin that could help your research, which would help me also, I am available. You may contact me.

      • Dear Farrell,

        Sorry but I am not related to the Crumley family. I was reading on this blog about your interesting family search going back to the 1700’s to learn about connecting to distant ancestors.

  4. For your information, I’ve learned that GEDMatch’s new Genesis affiliate allows database searchs of matching segments down to the 200 SNP and I believe 4 cM or lower thresholds, and one can specify the chromosome and beginning segment even – very useful for distant relative triangulation I believe.

      • I wonder if anyone would have time to comment on the following.

        I find the following 1-1 match on GEDMatch Genesis with me and an Englishman:
        Chr B37 Start Pos’n B37 End Pos’n Centimorgans (cM) SNPs
        13 30,877,359 33,307,358 5.8 2,337
        I also find the following 1-1 match with a possible English 8th cousin and the Englishman:
        Chr B37 Start Pos’n B37 End Pos’n Centimorgans (cM) SNPs
        13 31,608,949 33,567,055 3.9 2,166

        I also find the 1-1 match (not as large NSP) with a known 6th cousin 2x removed of mine (on same family branch as my possible 8th English cousin) and the 1st Englishman:

        Chr B37 Start Pos’n B37 End Pos’n Centimorgans (cM) SNPs
        13 26,017,374 27,045,129 2.6 255
        13 29,978,751 31,060,704 2.6 331
        13 31,143,391 32,429,355 3.3 383
        13 109,579,844 110,328,199 2.3 211

        I ‘m intrigued by the large number of SNP in the 1st two cases. Possible significance/evidence for common ancestor with possible 8th English cousin?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s