Go Back   Family Tree DNA Forums > Paternal Lineages (Y-DNA) > Paternal Lineage (Y-DNA STR) Advanced

Paternal Lineage (Y-DNA STR) Advanced Users can exchange information here on Y-DNA testing, results, and questions.

Reply
 
Thread Tools Display Modes
  #1  
Old 5th March 2018, 02:33 PM
FuriousGeorge FuriousGeorge is offline
FTDNA Customer
 
Join Date: May 2017
Posts: 28
Techniques for advanced STR comparison

Just curious if there's anything that can be done to distinguish people who match by coincidence from people who match because they are related, with any degree of certainty.

Are there any patterns to look for, such as smaller differences where there are differences, versus less differences of more steps?

I've googled a bit, but I'm not finding anything close to what I'm looking for. Maybe there's no way.
In the same vein, I'm having trouble finding a good source for which markers are compared in a step-wise versus infinite mutation model, after the 68th marker.
Reply With Quote
  #2  
Old 6th March 2018, 04:27 AM
FuriousGeorge FuriousGeorge is offline
FTDNA Customer
 
Join Date: May 2017
Posts: 28
One thing I'm doing is looking at places where I am above or below the modal value, and seeing if my suspected matches are on the same side.

I also have two markers that are considerably larger or smaller than the mode, but they are fast mutating, and I know that one is known for large 1-step mutations.

I'm also putting all of my possible results in a matrix over at http://www.mymcgee.com/tools/yutility111.html

I'm looking for patterns, such as someone who matches a lot of people more closely, and how that compares to how they match the modal value. There is some obvious clustering where people are related to each other. I have a person who I suspects matches me. I try to see if we both match other kits similarly, but I'm not sure if causality isn't running in the other direction.
Reply With Quote
  #3  
Old 8th March 2018, 03:45 PM
TwiddlingThumbs TwiddlingThumbs is offline
FTDNA Customer
 
Join Date: Jan 2016
Posts: 146
Quote:
Originally Posted by FuriousGeorge View Post
Just curious if there's anything that can be done to distinguish people who match by coincidence from people who match because they are related, with any degree of certainty.

Are there any patterns to look for, such as smaller differences where there are differences, versus less differences of more steps?

I've googled a bit, but I'm not finding anything close to what I'm looking for. Maybe there's no way.
Based solely on the STR results (without SNPs or genealogy), there is no way to know if the match is a result of a lack of mutations (a true close relationship) or convergent or back mutations (falsely showing a close relationship). Testing more markers can help, but you really need SNP testing or genealogy to really known.

Quote:
In the same vein, I'm having trouble finding a good source for which markers are compared in a step-wise versus infinite mutation model, after the 68th marker.
For how FTDNA calculates genetic distance, see the 4th paragraph at this link: http://www.ydnagroupingapp.com/stati...p_FAQ.html#how is gd calculated
Reply With Quote
  #4  
Old 8th March 2018, 04:11 PM
FuriousGeorge FuriousGeorge is offline
FTDNA Customer
 
Join Date: May 2017
Posts: 28
Yeah, I've noticed even identifying someone who is at least in the same parent clade as someone else, has less than a 1000 year TMRCA, then checking their markers manually, looking for something far from the mode that they match on and is a slow marker, /still/ does not guarantee you have found two people in the same terminal clade.

They are probably in sibling branches, though.

That said, if you know the age of the parent clade, you could probably adjust that linear math. It seems like it would be complex and multivariate, but you could take back mutations into account, essentially calculating how many could have happened given x years, and increase the GD accordingly, within a range.

Don't ask me for the #s though.
Reply With Quote
  #5  
Old 8th March 2018, 04:24 PM
TwiddlingThumbs TwiddlingThumbs is offline
FTDNA Customer
 
Join Date: Jan 2016
Posts: 146
Quote:
Originally Posted by FuriousGeorge View Post
Yeah, I've noticed even identifying someone who is at least in the same parent clade as someone else, has less than a 1000 year TMRCA, then checking their markers manually, looking for something far from the mode that they match on and is a slow marker, /still/ does not guarantee you have found two people in the same terminal clade.

They are probably in sibling branches, though.

That said, if you know the age of the parent clade, you could probably adjust that linear math. It seems like it would be complex and multivariate, but you could take back mutations into account, essentially calculating how many could have happened given x years, and increase the GD accordingly, within a range.

Don't ask me for the #s though.
I think some of the more sophisticated TMRCA algorithms do something like that, but you still just get a probabilistic answer with a very large confidence interval.
Reply With Quote
  #6  
Old 8th March 2018, 05:48 PM
dna dna is offline
FTDNA Customer
 
Join Date: Aug 2014
Posts: 2,963
Quote:
Originally Posted by TwiddlingThumbs View Post
Based solely on the STR results (without SNPs or genealogy), there is no way to know if the match is a result of a lack of mutations (a true close relationship) or convergent or back mutations (falsely showing a close relationship). Testing more markers can help, but you really need SNP testing or genealogy to really known. [----]
Amen.

Mr. W.
Reply With Quote
  #7  
Old 11th March 2018, 12:36 PM
wkauffman wkauffman is offline
FTDNA Customer
 
Join Date: Feb 2011
Posts: 157
SAPP

Look at http://www.jdvtools.com/SAPP/ but realize it could have some "time" constraints after which it may not work.
Reply With Quote
  #8  
Old 12th March 2018, 01:36 AM
FuriousGeorge FuriousGeorge is offline
FTDNA Customer
 
Join Date: May 2017
Posts: 28
Quote:
Originally Posted by wkauffman View Post
Look at http://www.jdvtools.com/SAPP/ but realize it could have some "time" constraints after which it may not work.
Cool. Thanks.

I'll see if the author would like to add my branch.
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

Similar Threads
Thread Thread Starter Forum Replies Last Post
X match comparison brian X-DNA Basics 1 27th October 2015 09:03 AM
GEDmatch Comparison mebrowndfw Family Finder Advanced Topics 21 25th August 2013 02:58 PM
Comparison Wills DNA and Genealogy for Beginners 25 3rd January 2013 10:53 AM
Family Finder Techniques to Share MarieRundquist Group Administrators - Advanced Chat 2 6th February 2012 07:48 AM
Recruiting Techniques Paul_Sheats Group Administrators - Advanced Chat 4 5th December 2006 08:59 AM


All times are GMT -5. The time now is 03:48 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.