EDIT (3/31/2015) – Beginning on April 1, 2015, I will no longer be able to accept submissions other than through the portal. My sincerest apologies, and I so greatly appreciate the files that have been submitted this way, but I have been inundated and won’t be able to take the extra time to process any non-portal submissions made after April 1, 2015. Thank you!
I need your help! I’m trying to gather data about the ranges of DNA shared by known relatives. How much DNA do you share with your sister? your brother? your second cousin? While it is possible to predict approximately how much DNA you share with a close relative, the actual numbers vary more than you might think.
If you’re interested in participating in this project, I’m looking for two numbers for the known relationship: (1) the total amount of shared DNA in cMs; and (2) the largest shared block in cMs. At Family Tree DNA, for example, you can find the numbers here:
You can contribute HERE! And if contribute before April 1, 2015 (and provide a valid email address), you will be eligible to win a FREE Family Finder kit! One free kit will be given away to a lucky winner.
This project will benefit all genetic genealogists; the aggregated information (minus emails of course) will be analyzed and put into a convenient user-friendly format that will be made available on the ISOGG wiki and here on the blog.
Thank you for participating!
.
Blaine,
Thank you collecting the DNA cM range values for different relationships. It will be a very valuable resource for all genetic genealogists!
Lisa
Thank you Lisa! I appreciate the response, it has been terrific.
But much more is needed, so keep them coming everyone!
If we are comparing at GEDmatch, what values do you want us to use for minimum segment size and SNP count?
Janet – Thank you so much for contributing! For this project, I only need the maximum cM and the largest segment. I’m not interested in the minimum segment size or SNPs. Thank you again!
Understood, but the numbers you showed from FTDNA would include all segments 1 cM and above, with at least 500 SNPs. The default values in the 1-to-1 compare at GEDmatch would only include segments 7 cM and 700 SNPs.
Very good point Janet, I’ll make a note for people to please use the default value at GEDmatch. One of the questions is what platform the data comes from (23andMe v. FTDNA v. GEDmatch), so I’ll be able to take the difference into consideration when performing the analysis. Thank you!
A great project! I submitted all mine from FTDNA, a few more may show up before April 1 or not, depending if that get those lab results done.
I have a few different first cousin & first once removed matches on 23andMe, but I may have to work harder to get the longest segment, as they are FIA.
Linda – thank you so much, it is greatly appreciated! Yes, it is sort of a pain to get this information from 23andMe, but much appreciated.
How far back do you care about the cousin marriages? I ask, because on my husband’s family he and siblings and uncles show higher than expected sharing — they did have a bunch of cousin marriages (darned Scots in the NC/TN/KY!) but it has been 4-5 generations since there were any. Same thing on my own family (Quakers for one and deep colonial New England for the other).
I also take it I can do this more than once (I administer several kits between the two families).
It’s going to be a bit subjective, but I wouldn’t worry too much about ancestors of 5 generations or more; the impact at that level will be very small. Feel free to make a brief note of it in the Notes field, if you think it could be important.
Thank you so much for taking the time!
Ok. I went through all my kits — without duplicating them I think! It was actually kind of fun to do! Hopefully this will help you.
Thank you very much Elizabeth!
As noted above, the FTDNA numbers include the small (mostly meaningless) segments, so they won’t be comparable to 23andMe. I think it’s still worth calibrating what to expect from FTDNA’s approach. I know the values for full siblings can cause some consternation, since FTDNA doesn’t give extra credit for completely identical segments and the shared cM drops well below the expected 50% level. Conversely, FTDNA’s values for various levels of cousins may turn out to be higher than “expected,” above and beyond the random nature of inheritance.
I agree Ann, and if I get enough data I’ll be able to do some delineation between 23andMe v. FTDNA v. GEDmatch. Based on a tiny handful of quick checks of my own data, I’ve estimated that the inclusion of segments below 7 cM at FTDNA will have an impact of less than 5% on the total shared DNA. I wonder if others see similar numbers?
Blaine, my numbers for differences between FTDNA vs GedMatch are somewhat higher than 5% (based on a very limited sampling).
For example, for a known 3rd cousin, twice removed relationship: FTDNA, total cM = 46, longest segment = 25 cM; same people using GedMatch, 40 cM & 30.3 cM, respectively.
Known fourth cousins, once removed: FTDNA 34 cM & 11 cM; GedMatch 30.6 & 30.6 cM.
Another fourth cousin, once removed: FTDNA 30 cM & 10 cM; GedMatch 9.7 & 9.7 cM.
I will provide what I can now have quite a few. I am waiting on 2 test to finish for my daughter and grand daughter then I will have 5 generations tested on one side and four on the other. Excited to see what the numbers look like between generation.
Thank you so much Lee Ann, it is greatly appreciated!
A worthy project. I will contribute.
Some questions:
1. If several close family members are available, are you interested in information for all of the relevant interconnections?
2. How are you defining endogamy?
Thank you Jason, much appreciated!
I am interested in all the relevant interconnections.
Regarding endogamy, in a way I’m leaving that up to the supplier of the information (most people who click “YES” also tend to provide a note in the note field to explain it). For me personally, I would define it as having any ancestry with a small founder population and/or intermarriage such that there is a known or suspected increase to the amount of DNA typically shared by known relatives.
Thank you again for contributions, it is highly appreciated.
For people who are on multiple platforms (some of my family are on 23andMe, FTDNA and GEDmatch), do you have a preference for which one we should enter? I assume you don’t want the sample relationship stats for 2 or more platforms, or do you?
Dave – that’s an interesting question. I will end up doing my analysis with results from each platform individually, I believe, with a little bit of cross-company analysis (not sure to what extent yet). So I’ll make a note of this, but feel free to submit stats for each platform.
I noticed that I was getting different numbers at 23andme vs familytree DNA for the same kits (V3 transferred) then I realized that the X is NOT included in family tree DNA so I summed them again. For my brother (smallest match is 7.4) I see 2606.87 at family tree DNA but the excel sheet gets a sum of 2670 for 23andme without the X, perhaps its a rounding issue?
Please note for my other 23andme kits I have to sort them by cM in the spreadsheet to get rid of the smaller than 7cM so perhaps you should mention that in your post.
I will try to mainly use the family tree DNA transferred kits for consistency
Kitty
Kitty,
The more I talk to others and the more I analyze the data, the more I realize that this will definitely require an analysis of company-by-company results. This will be really interesting!
Thank you again,
Blaine
For those with data from multiple companies, do you want results from each company? Would you accept a spreadsheet? I might have about 150 connections to report. Entering those into your form one by one would be much too time consuming.
Jason – since I will be doing analyses based on where the data comes from, I would indeed appreciate results from each company. I do accept spreadsheets.
Thank you again!
If we’re submitting spreadsheets, where do we send them? I have mine just about ready.
I have 29 results, where can I submit a spreadsheet? Also, all parent/child relationships at FTDNA have same length longest segment, presumably the X chromosome – is that what you want?
Karla and Duncan – please submit to [email protected]
Thank you!
That was interesting. Just finished entering several connection between myself and the fourteen closest family members who have tested. So interesting to see the range of amounts in common for the same type of relationships. In one instance my mom and her first cousin share more than twice as much as me and my first cousin. And the common grandparents that me and my cousin share were first cousins once removed. I would have though that we might share more than average, but we ended up sharing less than average. That recombination is a tricky business.
The kits I manage don’t have any 3C, but several 3C1R and 3C2R. Do you want those? I have sent the info for all the closer matches.
I have an interesting 3C 1R where the younger person is the earlier generation. Do you want it as well?
I’ve submitted the siblings and grandparents results that I currently have, and am unsure the other kit will be back and processed in time to assist with niece/nephew type results or not.
Looking forward to seeing the results!
Katy – feel free to submit the 3C1R. I don’t know if I’ll have sufficient submissions of 3C1R to do an analysis, but perhaps I will.
Thank you!
Thanks Blaine!
I don’t know how many 3C1R you’ll need, but there were at least 10 in the data I sent over. I simply included every instance of a known relationship, so that got me out to 3C2R and 4C.
It was kind of neat to see I had as much data to offer as I did, given that my family hasn’t yet connected with any previously unknown relatives who proved to be further out than fourth cousin.
I went ahead and added the 3C1R relationships that I had, too. I’m pretty sure I switched the total and longest on one or two of them.
So I decided to help out my cousins who have tested at 23andme by showing them how to get the numbers you want from those results in a blog post.
http://blog.kittycooper.com/2015/03/lets-all-help-collect-dna-statistics
Of course it is way easier to use GEDmatch!
Mr Bettinger, I sent you a hand full of my matches when you first started this project. I only sent my FTDNA ones I have actual paper trails to and actual proven relationships to. Do you also want our matches that don’t have paper trails too? Thanks so much for all you do.
Penny – thank you for your submissions! At this time I’m only looking for relationships with an established paper trail. Thank you, much appreciated!
Thanks for the quick reply. If I can prove up a few more of my matches before 1st April, I’ll send in the info. Thanks again.
Are you Penny that connects to the Craddocks of Cannock Wood ？
Maybe some explanation on how to get the information you want from 23andme? I can easily get % shared and number of segments, but I don’t see where to get largest segment and total?
Oh, I see Kitty’s comment above for 23andme. Several step process. 🙁
Hi Blaine, I just emailed you an Excel spreadsheet with 20 relationships among 9 people, including myself.
Thank you Cyndi!
I had ancestry and DNA tests done but have lost the results how can I retrieve the results thanks
Beginning on April 1, 2015, I will no longer be able to accept submissions other than through the portal. My sincerest apologies, and I so greatly appreciate the files that have been submitted this way, but I have been inundated and won’t be able to take the extra time to process any non-portal submissions made AFTER April 1, 2015. Thank you!
Are you still needing data points?
I just submitted 8 or 9 (I forget), and I am waiting on results from 2 additional tests that will provide 3 more sets from one kit and 7 from the other.
I too could supply more if needed. After all, I used the test I won to test my aunt!
I AM indeed still collection relationship data through the portal!
And how funny to provide the data you used the free kit to obtain! A self-serving award then, wasn’t it? Haha!
Great, I’ll pull that together and also try to think through which other known relatives have appeared since I submitted before.
You can also look out for this website which sells at-home based DNA kits which are best selling in today’s market. Take a look at it here. https://www.dnatestreview.org/