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 5th December 2017, 03:33 PM
Genealogic Genealogic is offline
FTDNA Customer
 
Join Date: Nov 2017
Posts: 4
Substantial errors in autosomal DNA tests?

These questions concerns both the FamilyTreeDNA tests, but more broadly also other autosomal tests.

1. FamilyTreeDNA uses a description of 24 population clusters at https://www.familytreedna.com/learn/...tion-clusters/

Concerning Finland this is at least partly erroneous, since it essentially only talks about the Finnic immigrations (tribes of the Finns, Tavastians, and Karelians, aside for the much earlier Sami).

Another, second migration of populations occurred the South-West were the Goths from 500-1000 A.C., particularly from Gotland and Sweden. Migration continued and accelerated from 1150 A.C. when most of the area of what today constitutes Finland became the eastern part of the kingdom of Sweden.

In fact, only about 62.5% of Finnish men belong to haplogroup N (“Finn”), while 28% belong to I1 (“Scandinavians”). Thus, if about one third of the population are Scandinavians, this key fact would seem fundamental to any discussion on what is meant by "Finns" and the people that populate the country today?


2.FamilyTreeDNA also states that "Genetically, however, influence from centuries of occupation and invasion is illustrated in the close genetic relatedness between populations in the Finland cluster and populations from Russia and Scandinavia.”

This is also partly untrue. While the genetic the influence from Scandinavia is substantial, the genetic influence from Russia has been quite minimal. Only about 5% of the population belong to R1a (“Slavs”). Moreover, the Scandinavian influence is substantially larger in the Western part of the country, particularly among the Swedish-speaking minority.

It would be nice to know from where FamilyTreeDNA gets the "close genetic relatedness between populations" in Russia? (The relatedness could plausibly be to the small Finnic-Uralic minorities living in Russia, but not with ~99% of Russians.)


3. The most serious problem with autosomal testing of the DNA seems to be on "population clusters" and ethnicities.

If we take the municipal of Larsmo in Finland, which is predominantly Swedish-speaking (~90%), we know from DNA testing that:
"The subpopulation LMO (Larsmo) differed significantly from all the other populations"... "The geographical substructure among the Finnish males was notable when measured with the ΦST values, reaching values as high as ΦST=0.227 in the Yfiler data. This is rather extreme, given that, e.g., subpopulations Larsmo and Kymi are separated by mere 400 km, with no apparent physical dispersal barriers between them".
Jukka U. Palo et al. 2008. The effect of number of loci on geographical structuring and forensic applicability of Y-STR data in Finland.
Int J Legal Med (2008)122:449-456

Moreover, another study showed that:

"Clear East-West duality was observed when the Finnish individuals were clustering using Geneland. Individuals from the Swedish-speaking part of Ostrobotnia clustered with Sweden when a joint analysis was performed on Swedish and Finnish autosomal genotypes".
Ulf Hannelius, 2008. Population genetic association and Zygosity testing on preamplified Dna. Karolinska Institutet. 2008

Thus, if the Swedish-speaking population in Ostrobothnia (where Larsmo is located) clustered with Sweden, it would be erroneous to assume the ancestors are “Finns” even if ~99% of all relatives today in Finland as seen in tests. I really do not see how this accounted for in the autosomal tests.


Of course, there are a number of other populations for which these errors would occur, e.g. French and Italian-speaking populations of Switzerland, and the German-speaking population of Italy. In all of these cases all the genetic cluster of these populations will be in the country they are born since the migration may have occurred over a thousand years ago.


It might be somewhat more tricky for a someone living in e.g. US to question and doubt the validity of autosomal DNA tests. However, my family has been active in genealogy for decades and I can now trace my ancestry in the Swedish-speaking part of Ostrobothnia hundreds of years. In some lines well over 30 generations and much further than the 1600-hundreds. (Compare this to autosomal tests, which begin becoming unreliable when you go back only 5 to 6 generations.)

In summary, we are talking of several hundreds (if not thousands) of people. A large majority of these have been born in Finland*). Perhaps 97% of these have typical Swedish names, while a few were from towns in Norway, Finland (with Finnish names), Denmark, Germany names (even one from Scotland).
*) Finland was actually the eastern part of Sweden for most of the time, i.e. from 1100 to 1809, and thus migration from Sweden was domestic.

According to an autosomal test my ancestry would be 52% Finnish. This is really exceptionally unlikely if

(1) the genealogical data do not show any more than perhaps 1% having a Finnish name. (Genealogic tree available if needed.)

(2) We also know from previous scientific studies that the region in which I trace my ancestry (in the Swedish-speaking Ostrobothnia, close to Larsmo) cluster with Sweden.

(3) About a third of all Finnish-speaking men actually belong to the Scandinavian haplogroup I1.

Thus, a more likely proposition to me is that the currently used interpretations of the autosomal DNA tests are invalid, particularly for minorities.
Reply With Quote
  #2  
Old 5th December 2017, 07:58 PM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,412
Quote:
Originally Posted by Genealogic View Post
These questions concerns both the FamilyTreeDNA tests, but more broadly also other autosomal tests.

1. FamilyTreeDNA uses a description of 24 population clusters at https://www.familytreedna.com/learn/...tion-clusters/

Concerning Finland this is at least partly erroneous, since it essentially only talks about the Finnic immigrations (tribes of the Finns, Tavastians, and Karelians, aside for the much earlier Sami).

Another, second migration of populations occurred the South-West were the Goths from 500-1000 A.C., particularly from Gotland and Sweden. Migration continued and accelerated from 1150 A.C. when most of the area of what today constitutes Finland became the eastern part of the kingdom of Sweden.

In fact, only about 62.5% of Finnish men belong to haplogroup N (“Finn”), while 28% belong to I1 (“Scandinavians”). Thus, if about one third of the population are Scandinavians, this key fact would seem fundamental to any discussion on what is meant by "Finns" and the people that populate the country today?


2.FamilyTreeDNA also states that "Genetically, however, influence from centuries of occupation and invasion is illustrated in the close genetic relatedness between populations in the Finland cluster and populations from Russia and Scandinavia.”

This is also partly untrue. While the genetic the influence from Scandinavia is substantial, the genetic influence from Russia has been quite minimal. Only about 5% of the population belong to R1a (“Slavs”). Moreover, the Scandinavian influence is substantially larger in the Western part of the country, particularly among the Swedish-speaking minority.

It would be nice to know from where FamilyTreeDNA gets the "close genetic relatedness between populations" in Russia? (The relatedness could plausibly be to the small Finnic-Uralic minorities living in Russia, but not with ~99% of Russians.)


3. The most serious problem with autosomal testing of the DNA seems to be on "population clusters" and ethnicities.

If we take the municipal of Larsmo in Finland, which is predominantly Swedish-speaking (~90%), we know from DNA testing that:
"The subpopulation LMO (Larsmo) differed significantly from all the other populations"... "The geographical substructure among the Finnish males was notable when measured with the ΦST values, reaching values as high as ΦST=0.227 in the Yfiler data. This is rather extreme, given that, e.g., subpopulations Larsmo and Kymi are separated by mere 400 km, with no apparent physical dispersal barriers between them".
Jukka U. Palo et al. 2008. The effect of number of loci on geographical structuring and forensic applicability of Y-STR data in Finland.
Int J Legal Med (2008)122:449-456

Moreover, another study showed that:

"Clear East-West duality was observed when the Finnish individuals were clustering using Geneland. Individuals from the Swedish-speaking part of Ostrobotnia clustered with Sweden when a joint analysis was performed on Swedish and Finnish autosomal genotypes".
Ulf Hannelius, 2008. Population genetic association and Zygosity testing on preamplified Dna. Karolinska Institutet. 2008

Thus, if the Swedish-speaking population in Ostrobothnia (where Larsmo is located) clustered with Sweden, it would be erroneous to assume the ancestors are “Finns” even if ~99% of all relatives today in Finland as seen in tests. I really do not see how this accounted for in the autosomal tests.


Of course, there are a number of other populations for which these errors would occur, e.g. French and Italian-speaking populations of Switzerland, and the German-speaking population of Italy. In all of these cases all the genetic cluster of these populations will be in the country they are born since the migration may have occurred over a thousand years ago.


It might be somewhat more tricky for a someone living in e.g. US to question and doubt the validity of autosomal DNA tests. However, my family has been active in genealogy for decades and I can now trace my ancestry in the Swedish-speaking part of Ostrobothnia hundreds of years. In some lines well over 30 generations and much further than the 1600-hundreds. (Compare this to autosomal tests, which begin becoming unreliable when you go back only 5 to 6 generations.)

In summary, we are talking of several hundreds (if not thousands) of people. A large majority of these have been born in Finland*). Perhaps 97% of these have typical Swedish names, while a few were from towns in Norway, Finland (with Finnish names), Denmark, Germany names (even one from Scotland).
*) Finland was actually the eastern part of Sweden for most of the time, i.e. from 1100 to 1809, and thus migration from Sweden was domestic.

According to an autosomal test my ancestry would be 52% Finnish. This is really exceptionally unlikely if

(1) the genealogical data do not show any more than perhaps 1% having a Finnish name. (Genealogic tree available if needed.)

(2) We also know from previous scientific studies that the region in which I trace my ancestry (in the Swedish-speaking Ostrobothnia, close to Larsmo) cluster with Sweden.

(3) About a third of all Finnish-speaking men actually belong to the Scandinavian haplogroup I1.

Thus, a more likely proposition to me is that the currently used interpretations of the autosomal DNA tests are invalid, particularly for minorities.
It should be understood, that anyone in any country usually shows a composite of regions of origin rather than have a single region of origin. I would not regard this as an error, although the picture is clearer at Gedmatch. (My daughter's maternal side is mostly Scandinavian-Norway, but my hunch is that she also has Sami lines)
Reply With Quote
  #3  
Old 5th December 2017, 09:22 PM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,412
Quote:
Originally Posted by josh w. View Post
It should be understood, that anyone in any country usually shows a composite of regions of origin rather than have a single region of origin. I would not regard this as an error, although the picture is clearer at Gedmatch. (My daughter's maternal side is mostly Scandinavian-Norway, but my hunch is that she also has Sami lines)
The main difference is that Gedmatch publishes the background information and Ftdna does not publish it. From the spreadsheet of the Gedmatch-Eurogenes calculator, Finns are: North Atlantic-34%, Baltic-48%, West Med-4% and Siberian-7%. Ftdna has a similar spreadsheet with different regions but it is kept private---Ftdna's regions are listed in My Origins.

Last edited by josh w.; 5th December 2017 at 09:29 PM.
Reply With Quote
  #4  
Old 6th December 2017, 08:48 AM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,412
Quote:
Originally Posted by josh w. View Post
The main difference is that Gedmatch publishes the background information and Ftdna does not publish it. From the spreadsheet of the Gedmatch-Eurogenes calculator, Finns are: North Atlantic-34%, Baltic-48%, West Med-4% and Siberian-7%. Ftdna has a similar spreadsheet with different regions but it is kept private---Ftdna's regions are listed in My Origins.
In other words, Ftdna has data showing that Finns are mainly a mixture of Baltic-Uralic and Scandinavian. Ftdna MO has regions which overlap the Gedmatch regions. Gedmatch also has an East Finn sample with a bit more Siberian and a bit less North Atlantic

In my own case I am 99% Ashkenazi at MO. Does that mean that Ftdna is unaware of my 50% European lines---even if they do not show the results.

Last edited by josh w.; 6th December 2017 at 09:01 AM.
Reply With Quote
  #5  
Old 6th December 2017, 12:00 PM
JMAisHere JMAisHere is offline
FTDNA Customer
 
Join Date: Oct 2017
Location: Sweden
Posts: 23
Grattis Finland 100 år !!!

Congratulations Finland, 100 years today.
Reply With Quote
  #6  
Old 6th December 2017, 12:57 PM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,412
Ftdna's MO spreadsheet

Quote:
Originally Posted by josh w. View Post
In other words, Ftdna has data showing that Finns are mainly a mixture of Baltic-Uralic and Scandinavian. Ftdna MO has regions which overlap the Gedmatch regions. Gedmatch also has an East Finn sample with a bit more Siberian and a bit less North Atlantic

In my own case I am 99% Ashkenazi at MO. Does that mean that Ftdna is unaware of my 50% European lines---even if they do not show the results.
My statement that Ftdna has conducted spreadsheet (or equivalent ) analysis is based on a Ftdna White Paper report. R Kahn stated that MO did not provide the full picture on Ashkenazi genetics because the spreadsheet type data would confuse or disappoint some customers
Reply With Quote
  #7  
Old 6th December 2017, 03:26 PM
Genealogic Genealogic is offline
FTDNA Customer
 
Join Date: Nov 2017
Posts: 4
The composite is not the error

Quote:
Originally Posted by josh w. View Post
It should be understood, that anyone in any country usually shows a composite of regions of origin rather than have a single region of origin. I would not regard this as an error, although the picture is clearer at Gedmatch. (My daughter's maternal side is mostly Scandinavian-Norway, but my hunch is that she also has Sami lines)
There seems to be a big misunderstanding here. The composite is not the error. The error is a composite of that is made up by the wrong ethnicities. A plausible reason might be because the interpretation of the DNA data is not sufficiently sophisticated - at least for some subpopulations and regions.

I just don't see the point of doing any autosomal testing if the validity is so poor.
Reply With Quote
  #8  
Old 6th December 2017, 04:14 PM
Genealogic Genealogic is offline
FTDNA Customer
 
Join Date: Nov 2017
Posts: 4
Two comments

Quote:
Originally Posted by josh w. View Post
The main difference is that Gedmatch publishes the background information and Ftdna does not publish it. From the spreadsheet of the Gedmatch-Eurogenes calculator, Finns are: North Atlantic-34%, Baltic-48%, West Med-4% and Siberian-7%. Ftdna has a similar spreadsheet with different regions but it is kept private---Ftdna's regions are listed in My Origins.
Two things here that I'd like to comment on. First, Finland shows a significant discrepancy of the ancestry between the east and west. The line goes from the south-east coast diagonally over Finland up to the north-west of the Gulf of Bothnia. West of the line I1 is much more prevalent, while in east the N-haplogroup. (This aside from the Lappland, were you also find Sami (U5b1b and V). Particularly among the Swedish-speaking population in Ostrobothnia, the Åland Island, Varsinais-suomi and in Satakunta. Any analysis of test results would of course need to reflect this.

Second, we know Finnish men (all populations and regions accumulated) consist of the haplogroups:
N - 61.5%
I1 - 28.0%
I2 - 0.5%
R1a - 5.0%
R1b - 3.5%
Others - 1.5%

Thus, it is debateable how the above groups of "North Atlantic-34%, Baltic-48%, West Med-4% and Siberian-7%." are defined.
Do the "Baltic" category refer to the Finnic ancestry of seen in haplogroup N? Why is there a substantial difference to the percentage of haplogroup N, which is 61.5%? And what is meant by "Siberian", is it R1a at 5%?
Reply With Quote
  #9  
Old 6th December 2017, 04:15 PM
Genealogic Genealogic is offline
FTDNA Customer
 
Join Date: Nov 2017
Posts: 4
Thumbs up

Quote:
Originally Posted by JMAisHere View Post
Congratulations Finland, 100 years today.
Reply With Quote
  #10  
Old 6th December 2017, 08:06 PM
josh w. josh w. is offline
FTDNA Customer
 
Join Date: Mar 2005
Posts: 2,412
Quote:
Originally Posted by Genealogic View Post
Two things here that I'd like to comment on. First, Finland shows a significant discrepancy of the ancestry between the east and west. The line goes from the south-east coast diagonally over Finland up to the north-west of the Gulf of Bothnia. West of the line I1 is much more prevalent, while in east the N-haplogroup. (This aside from the Lappland, were you also find Sami (U5b1b and V). Particularly among the Swedish-speaking population in Ostrobothnia, the Åland Island, Varsinais-suomi and in Satakunta. Any analysis of test results would of course need to reflect this.

Second, we know Finnish men (all populations and regions accumulated) consist of the haplogroups:
N - 61.5%
I1 - 28.0%
I2 - 0.5%
R1a - 5.0%
R1b - 3.5%
Others - 1.5%

Thus, it is debateable how the above groups of "North Atlantic-34%, Baltic-48%, West Med-4% and Siberian-7%." are defined.
Do the "Baltic" category refer to the Finnic ancestry of seen in haplogroup N? Why is there a substantial difference to the percentage of haplogroup N, which is 61.5%? And what is meant by "Siberian", is it R1a at 5%?
Don't confuse autosomal patterns with Y dna---they show some overlap but are not the same. Autosomal dna shows less of a gradient in Finland than Y dna. This does not mean that autosomal dna is in error. At Eurogenes there is no sample higher than 55% Baltic---it is not equivalent to any specific sample. Lithuania and Belarus are highest but East Finnish is close behind. If you don't like Ftdna or Eurogenes regions, try MDLP at Gedmatch.

Last edited by josh w.; 6th December 2017 at 08:10 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

Similar Threads
Thread Thread Starter Forum Replies Last Post
SOLVED: Autosomal Transfer result errors SurplusGadgets Features Requests & Bug Reports Area 0 28th September 2017 11:03 AM
A substantial prehistoric European ancestry amongst Ashkenazi maternal lineages girlperson1 Scientific Papers 17 22nd August 2014 07:27 AM
Autosomal STR and X-STR tests veramiller Family Finder Advanced Topics 2 15th January 2014 12:52 PM
New Autosomal Tests? Richard Rauch DNA and Genealogy for Beginners 23 17th September 2006 10:43 AM


All times are GMT -5. The time now is 09:34 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 - 2017, Jelsoft Enterprises Ltd.