Go Back   Family Tree DNA Forums > Universal Lineage Testing (Autosomal DNA) > Family Finder Advanced Topics

Family Finder Advanced Topics Advanced discussion about Family Tree DNA's Family Finder Product.

Reply
 
Thread Tools Display Modes
  #1  
Old 15th October 2011, 10:50 PM
shandy4473 shandy4473 is offline
FTDNA Customer
 
Join Date: Apr 2011
Posts: 394
Smile Examples of 3rd Cousin and 3rd Cousin first removed DNA Amounts

Good Day Everyone,

Would anyone happen to have any examples of what a 3rd Cousin and 3rd Cousin once removed DNA cM amounts would look like. I am looking for examples where people have gotten confirmed 3rd cousin matchings.

My dad and his mother matched to a person. Here are the amounts:

Mother - 65.13cm Longest Block - 42.96

Son - 62.02 Longest Block - 42.96

This is a carry over from a previous posting but I am curious to know of any confirmed 3rd cousins and their DNA cM amounts.
Reply With Quote
  #2  
Old 16th October 2011, 02:37 AM
rucksack rucksack is offline
Registered User
 
Join Date: Aug 2008
Posts: 640
Here is my own actual 3rd cousin, confirmed by DNA and paper compared with me:

Range / suggested / shared cM / longest block

2nd Cousin - 4th Cousin 3rd Cousin 107.34 23.48

Was that what you were looking for?
Reply With Quote
  #3  
Old 16th October 2011, 09:17 AM
shandy4473 shandy4473 is offline
FTDNA Customer
 
Join Date: Apr 2011
Posts: 394
Smile This is ok

Thanks rucksack,

This is what I am looking for. As you can the quantities are similar between my father and his mother which suggests the common ancestor is somewhat far back. FTDNA has the suggested relationship as a 3rd cousin for both my father and his mother. I could test other family members who could flesh out the exact relationship but I am wondering if I really need to????

Thanks
Steve
Reply With Quote
  #4  
Old 16th October 2011, 09:49 AM
shandy4473 shandy4473 is offline
FTDNA Customer
 
Join Date: Apr 2011
Posts: 394
Smile

Quote:
Originally Posted by shandy4473 View Post
Thanks rucksack,

This is what I am looking for. As you can see the quantities are similar between my father and his mother which suggests the common ancestor is somewhat far back. FTDNA has the suggested relationship as a 3rd cousin for both my father and his mother. I could test other family members who could flesh out the exact relationship but I am wondering if I really need to????

On the other hand - I am thinking that I really need to test other family members closet to my grandmother to reliably deduce the exact relationship

Thanks
Steve
Reply With Quote
  #5  
Old 16th October 2011, 11:12 AM
Brunetmj Brunetmj is offline
FTDNA Customer
 
Join Date: Jan 2011
Location: UpState NY
Posts: 1,053
Here is an example of a confirmed 3rd cousin one removed

Share cm 53.84
Longest block 24.65
Reply With Quote
  #6  
Old 17th October 2011, 12:01 AM
mkdexter mkdexter is offline
FTDNA Customer
 
Join Date: Dec 2009
Location: Mesa, AZ, USA
Posts: 3,021
Steve:

Other numbers for third cousins are simply not indicative of every situation like they would be for a first or second cousin scenario. The sum is too small beyond 2nd cousin.

It has to do with the algorithm. In fact as you can see in the results you posted, a parent and child have the same longest block, hence the same prediction, therefore unless you have this on paper you may not know the exact relation by these numbers alone. The same is true when comparing your numbers to the examples others might have.

Once you start seeing the same or similar numbers between a parent and their child compared to someone else, you know the segments are persisting over at least two generations and the algorithm will not know exactly which one is 3rd or 4th and which one is once or twice removed from that point. If it was a first or second cousin it would be easy, the one that has half the sum of the other one is the child and once removed.

Last edited by mkdexter; 17th October 2011 at 12:03 AM.
Reply With Quote
  #7  
Old 17th October 2011, 12:44 AM
shandy4473 shandy4473 is offline
FTDNA Customer
 
Join Date: Apr 2011
Posts: 394
Smile

Thanks Matt:

It looks like I am gonna have to ask some descendants of a relative of my grandmother to get the answers I am looking for. They are close enough to my grandmother's DNA profile to help flesh out and provide the answers I am looking for.

Steve
Reply With Quote
  #8  
Old 17th October 2011, 11:11 AM
Geneadict Geneadict is offline
Registered User
 
Join Date: Dec 2010
Posts: 328
@Matt: re: "it has to do with the algorithm"

Actually it has to to more so with the simple reality of autosomal recombination than it does with any algorithm that any company may attempt to apply.

While it appears that an actual relationship can be predicted within 1 degree of relationship out to 2nd cousin by either FF or 23andMe or anyone else using data from either company's Illumina SNP chip test, the wide range of possible results due to recombination makes this impossible to do for more distant relationships with any level of confidence.

Meanwhile the rule of "half" is just an "average" approximation and range of actual results may be higher or lower than this and have a high standard deviation. With each meiotic event separating the individuals the range continues to widen and by the time you get out to even 1st cousins to 1st cousins 1x removed, this range may overlap and thus the reason for the "within" one degree of relationship caveat above. By the time you are beyond 2nd cousin, this range is so wide and the average larger segment shares so small, it does not provide anyone ability to predict within that level of confidence and could possibly be many generations more distant.

Algorithms used to arrive at the predictions are primarily based on the largest segment, next largest segment and total share. However with total share including many segments under 5 cM may include several from different ancestors if an individuals overall ancestry is not diverse enough. For example, if you had a grandparent or great grandparent who was an actual 2nd or 3rd cousin to another grandparent or great grandparent (even on completely unconnected lines of your ancestry, maternal vs paternal), some of these smaller segments could easily persist and add to your overall share with someone descended from a cousin of the related parties in question.
Reply With Quote
  #9  
Old 17th October 2011, 11:53 AM
Javelin Javelin is offline
FTDNA Customer
 
Join Date: Jan 2009
Posts: 990
I don't have any third cousins proper tested, but you may wish to note the differences in longest blocks in some of these examples as well:

first cousin twice removed: 43.0
second cousin: 44.3 cM
second cousin once removed:44.5
third cousin once removed: 16.4 cM
third cousin once removed: 11.5 cM
Reply With Quote
  #10  
Old 17th October 2011, 12:10 PM
mkdexter mkdexter is offline
FTDNA Customer
 
Join Date: Dec 2009
Location: Mesa, AZ, USA
Posts: 3,021
Quote:
Originally Posted by Geneadict View Post
@Matt: re: "it has to do with the algorithm"

Actually it has to to more so with the simple reality of autosomal recombination than it does with any algorithm that any company may attempt to apply.

While it appears that an actual relationship can be predicted within 1 degree of relationship out to 2nd cousin by either FF or 23andMe or anyone else using data from either company's Illumina SNP chip test, the wide range of possible results due to recombination makes this impossible to do for more distant relationships with any level of confidence.

Meanwhile the rule of "half" is just an "average" approximation and range of actual results may be higher or lower than this and have a high standard deviation. With each meiotic event separating the individuals the range continues to widen and by the time you get out to even 1st cousins to 1st cousins 1x removed, this range may overlap and thus the reason for the "within" one degree of relationship caveat above. By the time you are beyond 2nd cousin, this range is so wide and the average larger segment shares so small, it does not provide anyone ability to predict within that level of confidence and could possibly be many generations more distant.

Algorithms used to arrive at the predictions are primarily based on the largest segment, next largest segment and total share. However with total share including many segments under 5 cM may include several from different ancestors if an individuals overall ancestry is not diverse enough. For example, if you had a grandparent or great grandparent who was an actual 2nd or 3rd cousin to another grandparent or great grandparent (even on completely unconnected lines of your ancestry, maternal vs paternal), some of these smaller segments could easily persist and add to your overall share with someone descended from a cousin of the related parties in question.
Paul that sounds real good but in the case (specifically) of FTDNA FF yes it has to do with their algorithm. Since the FF algorithm uses sum for 2nd cousin comparing to other 2nd cousin sums will give an approximation as to how they all compare. Since then FF algorithm uses longest block for 3rd cousin no comparisons to other tests can be mode. There is a definite difference in the operation that only FTDNA uses. If Steve was asking for other 2nd cousin results it would be different. In fact the best way to make a prediction for firsts, seconds, once and twice removed of people closer than 3rd cousins is by pattern recognition and not sum or longest block but no one does that (yet). Its something FTDNA is starting to get into as they use multiple longest block segments to make predictions.

My point to Steve is simply to not use sums or other people's results for a 3rd cousin comparisons.

Last edited by mkdexter; 17th October 2011 at 12:21 PM.
Reply With Quote
Reply

Bookmarks


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -5. The time now is 02:50 AM.


Family Tree DNA - World Headquarters

1445 North Loop West, Suite 820
Houston, Texas 77008, USA

Phone: (713) 868-1438 | Fax: (832) 201-7147
Copyright 2001-2010 Genealogy by Genetics, Ltd.
Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.