Am I Weird – Or What?

OK, don’t answer that.

When we’re working with our autosomal DNA and that of our close family members, we only see our immediate results.

For example, I can see that I share a total of 407.92cM segments of DNA with my second cousin, and our longest shared segment is 81.34cM.

2nd cousin share

Does that mean that another match with 407 shared segments is also a second cousin match?  In other words, is 407 normal for second cousins, or an outlier, and how would I know?

When we’re trying to figure out where a match might fit in the family – and how distantly related we really are to that person, we need a broader perspective.  We need to know what the average is, but we also need to know the possible range.  In other words, we need a lot more data than just our own to know where the most likely placement of this person is, and where they might fall in the range.  I mean, let’s face it, my family has not one time EVER fallen into the normal range of anything!

I have done a lot of cousin testing within the family, and I’ve compiled a chart of the amount of shared DNA that these various people share, along with the longest segment.  Why did I do this?

Blaine Bettinger is getting ready to do the grunt work. He’s going to crunch the data and prepare the ranges for how much DNA is shared by people of various relationships – and all he needs from us is raw data.

Blaine wrote about this recently, along with super-easy instructions for how to compile and contribute your data.  It’s right on your Family Finder match page – and Blaine makes submitting your information easy.

Blaine's instructions

You don’t need a whole list, like mine.  If there is an overachiever gene, I’m sure I have it.  I have contributed my information to Blaine already – because I think this kind of a project – to be shared among all genetic genealogists, is very important and will benefit everyone. It takes very little on our part and a whole lot more work on Blaine’s part – and I’m not only very grateful – I’m more than willing to submit my data.  And hey, if you submit your data before April 1st, you are eligible for a Family Finder kit from Blaine – and no – I don’t think that’s an April Fool’s joke.  Maybe you’ll get a kit with no swabs:)

So, how weird am I anyway?

I sorted my spreadsheet by the total shared DNA segments and I added a column for the percentage of expected shared DNA.

You can see that this holds together quite well and just as expected through the 1st cousin/great-granddaughter range which is 12.5% expected shared DNA.

Amount shared DNA

However, below that, the expected amount of DNA and the actual amount of DNA becomes shuffled or “out of order.”  For example you can see that I have a second cousin once removed that I share 460cM of DNA with – and I should share 1.56% of DNA with them.  Just below that row, I have a first cousin once removed with whom I share 414cM of DNA with, and who I’m supposed to share about 6.25% of DNA.

We’ve talked several times about how DNA isn’t always passed in 50% amounts, even though that is the average, and this shows the cumulative effects of this phenomenon, especially after you’re out a couple of generations.

So, now we wait. I want to know if my family falls into the middle or expected range.  I want to know if my family is an outlier.  Maybe, for a change, we’re normal.  Nah!!!  I want to know just how weird we really are.

Don’t answer that….

Just send your data to Blaine and he’ll take care of that answer for all of us.

33 thoughts on “Am I Weird – Or What?

  1. Roberta I really enjoy reading your emails even though I am sure I don’t understand even 20% of the content. I am glad there are folks like you trying to educate the ‘great unwashed’ like me.

    Martin Smith

  2. Well duh, and double duh……….I know what a 1st cousin 1R is because I have one, and he is the son of my 1st cousin. But, I thought a 1st cousin 1R was EXACTLY the same,( but just another way of saying it ) as a 2nd cousin; but obviously I am wrong because although they share about the same cMs for longest segment, they have a big difference in Total Shared cMs.

    So, who qualifies as a 2nd Cousin?

    • Draw a chart. Start with your great-grandparents and draw three generation below two of their children.

      Their children are siblings.

      The next generation are full first cousins to each other.

      The next generation are full second cousins to each other.

      But if you have one person from the first cousin level and one person from the second cousin level, they are first cousins once removed from each other.

  3. Thanks for the heads up. I gladly sent some info to Mr. Bettinger. Can’t wait to see the results of his project.

  4. Mathematically you are correct, but the issue is not the probability of inheriting segments so much as the likelihood of inter-related ancestors. Small towns and villages tend to shuffle the deck every few generations so to speak, And the issue of matching is compounded by the same occurrence in the person you are matching. So the shared cMs are completely misleading, other than to alert you to the fact that multiple relationship paths are likely present. I suspect this is much more prevalent than is commonly believed, in areas that did not have much in migration. And then triangulation becomes problematic.
    Keep the posts coming …. they are very enlightening, and anything to help deal with multiple matching lines would be welcome.

  5. I just sent in my brothers and mine, we have different numbers between 23&me and FTDNA.
    FTDNA= 2746.49…214.28. 23&me= 2893…223. I’ll have to wait until more of the family tests.

    That one adoptee that I match, but haven’t found his parents yet, 20412…95cm to me. 5380…..35.1 to my brother. He only matches me on ch6, but matches Rick on chs 6, 9 & 20.

    So far he is the closest match on both sites, I don’t think he has done FTDNA, just 23&me.

    I love all this info Cousin!

    Cathee

  6. I just sent my contributions. It will be interesting to see what results he gets. I have a 3rd cousin that I don’t share any DNA with and I was baffled by this finding. I am wondering if I should send 0 cM for a 3rd cousin to Blaine. I don’t want to skew the results however.

    • Yes, you should. Not sending it will skew the results:) If you and that cousin are both at Gedmatch, try running the comparison there. You likely do share DNA, just not over the vendor threshold elsewhere.

      • OK. I just sent in my result of 0 cM for my 3rd cousin. The 0 cM comes from Gedmatch. We share the same Great Great Grandparents and we both share > 60 cM with another cousin who descends from the same couple.

      • There is a difference between “not sharing any DNA” and “not showing up as a match with the testing company.” If you are going to start reporting empty results, you must differentiate between them.

      • Israel – You’re 100% correct, but for purposes of this study I’ll have to lump the two categories together, unfortunately. I’m doing platform-based analyses (FTDNA, 23andMe, GEDmatch, etc.), and asking for the GEDmatch default, so this project can’t decipher between “not sharing any DNA” and “not showing up as a match with the testing company”. Any report that issues from this project will have to note that.

        Luckily, this will only affect a small number of the submissions, beyond the second cousin range (and I’m only going to the third cousin range at most anyway), since I don’t believe anyone has proven a second cousin match that failed to share DNA.

      • Oh, I went to the ISOGG autosomal DNA statistics page and utilized the percentages there. If you’re asking how to calculate the actual percentage, take the total cm of 6676cm and divide into the total amount shared, like 700cM. That would be 700 divided by 7776 which is .1049 or about 10%.

  7. Are the numbers shared, always exact, that is in agreement for every testing company?
    My son shares with me via 23andme 47.2% ( 3552.3 total cm and 214.5 is longest cm ), he shares with his mother ( 3587.1 total cm and 281.5 is longest cm ).
    If I use my ftdna data with my son in gedmatch it becomes 48.6%. I am unsure what is going on or what numbers are lost.
    Are these the sort of numbers that Blaine requires?

    victor

    • Victor – the numbers will vary from company to company based on a variety of factors. Once such factor, for example, is whether the X is included in the total cM count.

      You can submit FTDNA, 23andMe, and GEDmatch data since I will be doing analyses by platform. I’m looking for data in cM form (total cM and longest cM block).

      Thank you,
      Blaine

  8. I can probably give Blaine 230 matches, but using his form one at a time is alot of work.

    If I can do it the way you have it, Roberta, as a chart, I am more likely to participate.

    And don’t forget to tell him when people are known relatives multiple ways, so he can either take that into account or ignore that entry.

    • Israel – my apologies, I have been meaning to get back to you, but this project has absolutely buried me in email! If you could provide a chart like the one above (in Word, Excel, or some similar form), that would be so very convenient for me!

  9. Roberta,
    I don’t know how I missed this post when it first came out. This is something I’ve been struggling with for months, now. I think my autosomal family tree must be in the weird category, also. We clearly don’t fit the expected norms. However, I believe that is to be expected. As I’ve struggled to make sense of the many “large segment but unknown connection matches,” I’ve tried numerous figures of merit to try and match genetic distance and one or more parameters related to the shared dna. I believe sanity may only lie in looking at a larger sample size, as you noted in your blog. So, I look forward to the results of Blaine’s number crunching. I’ve noticed two factors that seem to skew my data, but clearly are not the entire story: 1) multiple sets of common ancestors with a particular match, and 2) having one or more ancestral generations where close cousins married. This seems to amplify both the amount of shared dna and possibly (but not clearly) the longest segment size.

    Thank you for a very timely and interesting article!

    Ray Baker

  10. Pingback: DNAeXplain Archives – General Information Articles | DNAeXplained – Genetic Genealogy

  11. Pingback: The Best and Worst of 2015 – Genetic Genealogy Year in Review | DNAeXplained – Genetic Genealogy

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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