Page 1 of 2 12 LastLast
Results 1 to 10 of 13

Thread: What’s the difference between distance and single on G25 Vahaduo?

  1. #1
    Registered Users
    Posts
    194
    Sex

    What’s the difference between distance and single on G25 Vahaduo?

    Which is the best to use?

  2. The Following 3 Users Say Thank You to HBA39 For This Useful Post:

     JMcB (04-21-2021),  Nino90 (04-21-2021),  Trelvern (04-21-2021)

  3. #2
    Registered Users
    Posts
    2,444
    Sex
    Location
    Sweden
    Ethnicity
    Italo-Germanic
    Nationality
    Swedish
    Y-DNA (P)
    R-L2 / R1b-U152
    mtDNA (M)
    H1a1

    Sweden Italy Italy 1861-1946 Sami Vatican Italy Tuscany
    Quote Originally Posted by HBA39 View Post
    Which is the best to use?
    Depens on what you want to show.

    Distance. Populations or samples who are most close to your dna.

    Single is like a oracle. It will break down the components to the possible best fit.
    That does not mean it shows your real ancestry. But possible combinations rather.


    I take myself as an example,
    Below is the modern average populations with closest distance to myself:



    Distance to: Nino_scaled
     
    0.03392686 Afrikaner
    0.03546667 Dutch
    0.03672958 German
    0.03711056 French_Brittany
    0.03760816 Welsh
    0.03796138 Norwegian
    0.03801978 French_Pas-de-Calais
    0.03904314 Belgian
    0.03951269 English
    0.04001975 Danish
    0.04010249 Austrian
    0.04041384 Scottish
    0.04093370 French_Nord
    0.04134256 Orcadian
    0.04136284 Irish
    0.04159342 English_Cornwall
    0.04178019 Swedish
    0.04194514 Shetlandic
    0.04313361 French_Alsace
    0.04353401 French_Paris
    0.04412031 Icelandic
    0.04566962 Swiss_German
    0.04663781 French_Seine-Maritime
    0.04808226 Hungarian
    0.04935593 Czech


    As you see. I am most close to Afrikander ( White South Afrikans ) and dutch, Germans etc.
    This does not mean I am Dutch or Afrikander. I am not. But i have similar levels of ancient components as these peoples.

    If you are not of mixed European ancestry. These calculators will work better most of the times.



    Here is my "Single" :

    Target: Nino_scaled
    Distance: 2.1962% / 0.02196205 | ADC: 0.25x RC
    44.4 Norwegian
    24.4 Spanish_Soria
    12.0 Swiss_French
    11.6 Cossack_Kuban
    5.2 Slovakian
    1.4 Nganassan
    1.0 Itelmen
    Shows Scandinavian admixture (I am around 50-65% Swedish)
    and also Spanish + Swiss French (Possible proxy for my Italian ancestry).

    Cossack_Kuban (I have no eastern european ancestry, this component is somewhat shown often in Swedes. But it is not real).

    Nganssan + Itelmen = Show my Siberian admixture from my Saami side.


    If you choose "REDUCE" you can set the populations from 3 to 7.

    Here is my results when choosing 3:

    Target: Nino_scaled
    Distance: 2.3236% / 0.02323550 | R3P
    60.6 Swedish
    36.2 Spanish_Soria
    3.2 Itelmen
    Last edited by Nino90; 04-21-2021 at 12:46 PM.
    Hidden Content

    Target: Nino_scaled
    Distance: 3.1640% / 0.03163963 | R3P
    64.4 __Germanic
    26.4 __IllyricRoman
    9.2 __Finnic

  4. The Following 3 Users Say Thank You to Nino90 For This Useful Post:

     HBA39 (04-21-2021),  JMcB (04-21-2021),  Trelvern (04-21-2021)

  5. #3
    Registered Users
    Posts
    194
    Sex

    Quote Originally Posted by Nino90 View Post
    Depens on what you want to show.

    Distance. Populations or samples who are most close to your dna.

    Single is like a oracle. It will break down the components to the possible best fit.
    That does not mean it shows your real ancestry. But possible combinations rather.


    I take myself as an example,
    Below is the modern average populations with closest distance to myself:



    Distance to: Nino_scaled
     
    0.03392686 Afrikaner
    0.03546667 Dutch
    0.03672958 German
    0.03711056 French_Brittany
    0.03760816 Welsh
    0.03796138 Norwegian
    0.03801978 French_Pas-de-Calais
    0.03904314 Belgian
    0.03951269 English
    0.04001975 Danish
    0.04010249 Austrian
    0.04041384 Scottish
    0.04093370 French_Nord
    0.04134256 Orcadian
    0.04136284 Irish
    0.04159342 English_Cornwall
    0.04178019 Swedish
    0.04194514 Shetlandic
    0.04313361 French_Alsace
    0.04353401 French_Paris
    0.04412031 Icelandic
    0.04566962 Swiss_German
    0.04663781 French_Seine-Maritime
    0.04808226 Hungarian
    0.04935593 Czech


    As you see. I am most close to Afrikander ( White South Afrikans ) and dutch, Germans etc.
    This does not mean I am Dutch or Afrikander. I am not. But i have similar levels of ancient components as these peoples.

    If you are not of mixed European ancestry. These calculators will work better most of the times.



    Here is my "Single" :



    Shows Scandinavian admixture (I am around 50-65% Swedish)
    and also Spanish + Swiss French (Possible proxy for my Italian ancestry).

    Cossack_Kuban (I have no eastern european ancestry, this component is somewhat shown often in Swedes. But it is not real).

    Nganssan + Itelmen = Show my Siberian admixture from my Saami side.


    If you choose "REDUCE" you can set the populations from 3 to 7.

    Here is my results when choosing 3:
    Thanks so much. When I use single mode the report isn’t even close to my known ancestry. It leaves out large chunks of Eastern Europe and focuses on Northern Europe. But with the distance tool it’s focused more on my Eastern Europe ancestry rather than Northern.

  6. The Following 2 Users Say Thank You to HBA39 For This Useful Post:

     JMcB (04-21-2021),  Nino90 (04-22-2021)

  7. #4
    Gold Class Member
    Posts
    376
    Sex
    Y-DNA (P)
    R-Y33
    mtDNA (M)
    J1c2
    Y-DNA (M)
    E-Y6938

    Russian Federation Russia Tatarstan Arms Russia Mordovia Timurid
    G25 values are coordinates on a Principal Component Analysis (PCA) plot. They measure the genetic distance between groups, so when you're dealing with an individual with ancestry from two populations - nMonte will often pick the group placed somewhere between the two true source populations on that PCA plot over the true sources.

    An offhand example is somebody who's half British/Sicilian will often cluster near Northern Italians on the PCA. So on some nMonte runs will get 100% Northern Italian rather than the 50%-50% British/Sicilian split.

    From personal experience, playing around with the penalty distance and reduce the population option will get you better results.

    Here the first run is with a 0.25 penalty, and the second run is without the penalty but with a reduction of possible source populations to three.

     

    Target: Alex_T
    Distance: 0.4858% / 0.00485833 | ADC: 0.25x RC
    29.0 Mordovian
    23.8 Moldovan
    16.8 Tatar_Mishar
    11.6 Turkish_Deliorman
    11.4 Turkish_Rumeli
    4.4 Cossack_Kuban
    3.0 Berber_Tunisia_Chen


     

    Target: Alex_T
    Distance: 0.3437% / 0.00343723 | R3P
    60.0 Ukrainian
    24.6 Ashkenazi_Russia
    15.4 Bashkir


    The second run is considerably more accurate but still far more perfect in terms of what I know about my ancestry.
    YDNA (P): R-Y33
    YDNA (P, maternal line): R-Y20756
    YDNA(M): E-Y6938

  8. The Following 5 Users Say Thank You to altvred For This Useful Post:

     CyrylBojarski (04-21-2021),  HBA39 (04-22-2021),  JerryS. (05-08-2021),  JMcB (04-21-2021),  Nino90 (04-22-2021)

  9. #5
    Registered Users
    Posts
    194
    Sex

    Quote Originally Posted by altvred View Post
    G25 values are coordinates on a Principal Component Analysis (PCA) plot. They measure the genetic distance between groups, so when you're dealing with an individual with ancestry from two populations - nMonte will often pick the group placed somewhere between the two true source populations on that PCA plot over the true sources.

    An offhand example is somebody who's half British/Sicilian will often cluster near Northern Italians on the PCA. So on some nMonte runs will get 100% Northern Italian rather than the 50%-50% British/Sicilian split.

    From personal experience, playing around with the penalty distance and reduce the population option will get you better results.

    Here the first run is with a 0.25 penalty, and the second run is without the penalty but with a reduction of possible source populations to three.

     

    Target: Alex_T
    Distance: 0.4858% / 0.00485833 | ADC: 0.25x RC
    29.0 Mordovian
    23.8 Moldovan
    16.8 Tatar_Mishar
    11.6 Turkish_Deliorman
    11.4 Turkish_Rumeli
    4.4 Cossack_Kuban
    3.0 Berber_Tunisia_Chen


     

    Target: Alex_T
    Distance: 0.3437% / 0.00343723 | R3P
    60.0 Ukrainian
    24.6 Ashkenazi_Russia
    15.4 Bashkir


    The second run is considerably more accurate but still far more perfect in terms of what I know about my ancestry.

    Thanks a lot. Somebody send me a link to PCA plot it it shows that I plot between Romance and Slavic with that tool. Which is a bit more accurate. I’m a mix of Northern Europe, Eastern Europe and West Asian so I guess that’s why it’s harder.

  10. The Following 2 Users Say Thank You to HBA39 For This Useful Post:

     JMcB (04-21-2021),  Nino90 (04-22-2021)

  11. #6
    Registered Users
    Posts
    1,891
    Sex
    Location
    Brittany
    Nationality
    French
    Y-DNA (P)
    R-U106>DF98>BY650
    mtDNA (M)
    K1a

    France Bretagne Trégor France Bretagne Kroaz Du
    Same calculations ( my known ancestry is very simple : 100% Breton ie Brittany France) Usually tests gave me : c 80% NW France Britain and Ireland and c 20% Spain where no ancestry at all (due to Early Farmers i guess).

    So:

    Distance to: Trelvern_scaled

    0.01919055 French_Brittany
    0.02229306 Welsh
    0.02312409 English_Cornwall
    0.02315099 Belgian
    0.02420877 French_Nord
    0.02461122 English
    0.02629482 Scottish
    0.02704943 French_Paris
    0.02771466 Orcadian
    0.02803256 Irish
    0.02812430 French_Pas-de-Calais
    0.02936130 French_Alsace
    0.03003258 Afrikaner
    0.03064985 Swiss_German
    0.03068600 Dutch
    0.03085576 German
    0.03092048 French_Seine-Maritime
    0.03161421 Shetlandic
    0.03429936 French_Occitanie
    0.03553250 Norwegian
    0.03618271 Danish
    0.03789352 French_Auvergne
    0.03793997 Icelandic
    0.04104319 Austrian
    0.04112135 Swiss_French




    Target: Trelvern_scaled
    Distance: 1.5262% / 0.01526160 | ADC: 0.25x RC

    57.8 Welsh
    21.4 Spanish_Navarra
    17.0 English_Cornwall
    2.4 Kaitag
    0.8 Ket
    0.6 Wichi

    No ancestry in Wales or Cornwall but similarities. Atlantic NW ancestry is 57.8+17 ans Atlantic SW is 21.4 . I think it's reliable .




    Target: Trelvern_scaled
    Distance: 1.5627% / 0.01562695 | R3P

    54.0 Norwegian
    35.6 Basque_Spanish
    10.4 Kaitag


    Norwegian is a proxy for NW ancestry. (MyHeritage gave me about 30% of Scandinavian ancestry but it's a joke)
    Last edited by Trelvern; 04-21-2021 at 03:01 PM.

  12. The Following 2 Users Say Thank You to Trelvern For This Useful Post:

     JMcB (04-21-2021),  Nino90 (04-22-2021)

  13. #7
    Registered Users
    Posts
    194
    Sex

    Quote Originally Posted by Trelvern View Post
    Same calculations ( my known ancestry is very simple : 100% Breton ie Brittany France) Usually tests gave me : c 80% NW France Britain and Ireland and c 20% Spain where no ancestry at all (due to Early Farmers i guess).

    So:

    Distance to: Trelvern_scaled

    0.01919055 French_Brittany
    0.02229306 Welsh
    0.02312409 English_Cornwall
    0.02315099 Belgian
    0.02420877 French_Nord
    0.02461122 English
    0.02629482 Scottish
    0.02704943 French_Paris
    0.02771466 Orcadian
    0.02803256 Irish
    0.02812430 French_Pas-de-Calais
    0.02936130 French_Alsace
    0.03003258 Afrikaner
    0.03064985 Swiss_German
    0.03068600 Dutch
    0.03085576 German
    0.03092048 French_Seine-Maritime
    0.03161421 Shetlandic
    0.03429936 French_Occitanie
    0.03553250 Norwegian
    0.03618271 Danish
    0.03789352 French_Auvergne
    0.03793997 Icelandic
    0.04104319 Austrian
    0.04112135 Swiss_French




    Target: Trelvern_scaled
    Distance: 1.5262% / 0.01526160 | ADC: 0.25x RC

    57.8 Welsh
    21.4 Spanish_Navarra
    17.0 English_Cornwall
    2.4 Kaitag
    0.8 Ket
    0.6 Wichi

    No ancestry in Wales or Cornwall but similarities.



    Target: Trelvern_scaled
    Distance: 1.5627% / 0.01562695 | R3P

    54.0 Norwegian
    35.6 Basque_Spanish
    10.4 Kaitag


    Norwegian is a proxy for NW ancestry. (MyHeritage gave me about 30% of Scandinavian ancestry but it's a joke)

    Yours look similar to mine with single mode it’s a lot more northern shifted but I’m only around 30% northern in reality where as it’s estimating me to be around 50-80%, except the distance mode easily picks up on Balkan and Eastern Europe ancestry. I’m kinda bummed it doesn’t show up in single mode.

  14. The Following User Says Thank You to HBA39 For This Useful Post:

     Trelvern (04-21-2021)

  15. #8
    Registered Users
    Posts
    194
    Sex

    Quote Originally Posted by Trelvern View Post
    Same calculations ( my known ancestry is very simple : 100% Breton ie Brittany France) Usually tests gave me : c 80% NW France Britain and Ireland and c 20% Spain where no ancestry at all (due to Early Farmers i guess).

    So:

    Distance to: Trelvern_scaled

    0.01919055 French_Brittany
    0.02229306 Welsh
    0.02312409 English_Cornwall
    0.02315099 Belgian
    0.02420877 French_Nord
    0.02461122 English
    0.02629482 Scottish
    0.02704943 French_Paris
    0.02771466 Orcadian
    0.02803256 Irish
    0.02812430 French_Pas-de-Calais
    0.02936130 French_Alsace
    0.03003258 Afrikaner
    0.03064985 Swiss_German
    0.03068600 Dutch
    0.03085576 German
    0.03092048 French_Seine-Maritime
    0.03161421 Shetlandic
    0.03429936 French_Occitanie
    0.03553250 Norwegian
    0.03618271 Danish
    0.03789352 French_Auvergne
    0.03793997 Icelandic
    0.04104319 Austrian
    0.04112135 Swiss_French




    Target: Trelvern_scaled
    Distance: 1.5262% / 0.01526160 | ADC: 0.25x RC

    57.8 Welsh
    21.4 Spanish_Navarra
    17.0 English_Cornwall
    2.4 Kaitag
    0.8 Ket
    0.6 Wichi

    No ancestry in Wales or Cornwall but similarities. Atlantic NW ancestry is 57.8+17 ans Atlantic SW is 21.4 . I think it's accurate.



    Target: Trelvern_scaled
    Distance: 1.5627% / 0.01562695 | R3P

    54.0 Norwegian
    35.6 Basque_Spanish
    10.4 Kaitag


    Norwegian is a proxy for NW ancestry. (MyHeritage gave me about 30% of Scandinavian ancestry but it's a joke)


    Sorry for the big post but here are a few of my distances for example
    G25 Modern individual samples scaled
    https://ibb.co/c26LdQ7

    Modern Europe G25.20
    https://ibb.co/YQcqYH9

    SouthEast Europe calculator
    https://ibb.co/qsP9dns

    UN Geographical world regions
    https://ibb.co/CKNT5zc

    Eurogenes K36 like G25 model
    https://ibb.co/1ZnySwr

    K53
    https://ibb.co/9b5h7vc

    K64
    https://ibb.co/R6G3xS4

    GenCove like calculator
    https://ibb.co/qyd0F17

    European and MENA calculator
    https://ibb.co/SnvF7Gx

    K35 Modern calculator
    https://ibb.co/TTY8DKR

    K25 calculator
    https://ibb.co/12gCXnh

    AncestryDNA model
    https://ibb.co/djB3wCJ

    LivingDNA model
    https://ibb.co/3zdhKQw

    MyHeritage model
    https://ibb.co/5smBfvL

    G25 language
    https://ibb.co/C72mVGQ

    Modern K64
    https://ibb.co/P651zRX

    Supreme regional calculator
    https://ibb.co/r0kJfwR

    World 134
    https://ibb.co/2ZZmrnL

    Supreme world k39
    https://ibb.co/YjnZpGX

    G25 calculator
    https://ibb.co/GHHJNjy

    West Eurasian calculator
    https://ibb.co/LRMcXSt

  16. The Following 2 Users Say Thank You to HBA39 For This Useful Post:

     Nino90 (04-30-2021),  Trelvern (04-22-2021)

  17. #9
    Registered Users
    Posts
    194
    Sex

    Quote Originally Posted by Nino90 View Post
    Depens on what you want to show.

    Distance. Populations or samples who are most close to your dna.

    Single is like a oracle. It will break down the components to the possible best fit.
    That does not mean it shows your real ancestry. But possible combinations rather.


    I take myself as an example,
    Below is the modern average populations with closest distance to myself:



    Distance to: Nino_scaled
     
    0.03392686 Afrikaner
    0.03546667 Dutch
    0.03672958 German
    0.03711056 French_Brittany
    0.03760816 Welsh
    0.03796138 Norwegian
    0.03801978 French_Pas-de-Calais
    0.03904314 Belgian
    0.03951269 English
    0.04001975 Danish
    0.04010249 Austrian
    0.04041384 Scottish
    0.04093370 French_Nord
    0.04134256 Orcadian
    0.04136284 Irish
    0.04159342 English_Cornwall
    0.04178019 Swedish
    0.04194514 Shetlandic
    0.04313361 French_Alsace
    0.04353401 French_Paris
    0.04412031 Icelandic
    0.04566962 Swiss_German
    0.04663781 French_Seine-Maritime
    0.04808226 Hungarian
    0.04935593 Czech


    As you see. I am most close to Afrikander ( White South Afrikans ) and dutch, Germans etc.
    This does not mean I am Dutch or Afrikander. I am not. But i have similar levels of ancient components as these peoples.

    If you are not of mixed European ancestry. These calculators will work better most of the times.



    Here is my "Single" :



    Shows Scandinavian admixture (I am around 50-65% Swedish)
    and also Spanish + Swiss French (Possible proxy for my Italian ancestry).

    Cossack_Kuban (I have no eastern european ancestry, this component is somewhat shown often in Swedes. But it is not real).

    Nganssan + Itelmen = Show my Siberian admixture from my Saami side.


    If you choose "REDUCE" you can set the populations from 3 to 7.

    Here is my results when choosing 3:


    Would you mind helping me analyse these results please?

    https://ibb.co/9vh7MR6
    https://ibb.co/LzyPgDd
    https://ibb.co/TP4RswP
    https://ibb.co/86MjXkk
    https://ibb.co/0VNwq8h
    https://ibb.co/1vh3HgX
    https://ibb.co/wYzZH33
    https://ibb.co/4TtSTr2
    https://ibb.co/RBtQRwb
    https://ibb.co/pxykDG9

  18. #10
    Registered Users
    Posts
    1,101
    Sex
    Y-DNA (P)
    R1a1a1b1a2b3
    mtDNA (M)
    U5a2b

    Polish–LithuanianCommonwealth Bulgaria Byzantine Empire Russian Federation Israel Russia Imperial
    Quote Originally Posted by altvred View Post
    The second run is considerably more accurate but still far more perfect in terms of what I know about my ancestry.
    I think second run is very similiar to what you get on 23andme.

    In my case only deleting majority of samples can improve results. Even reducing populations to 3 with all samples cant

    Here is what I get with all samples on 0.25x

     
    Target: Cyryl_scaled
    Distance: 1.6759% / 0.01675899 | ADC: 0.25x RC
    49.0 Ukrainian
    21.8 Bosnian
    19.8 Sorb_Niederlausitz
    5.8 Mountain_Jew
    2.4 Mari
    1.2 German_East



    Only Polish, Balkan, Western Asian samples

     

    Target: Cyryl_scaled
    Distance: 1.8847% / 0.01884716
    70.2 Polish_Southeast
    15.8 Polish_West
    7.2 Greek_Trabzon
    6.8 Albanian


    Relatively good results in 2 run, but still only regionally haha, even if you are 1/8 something different from majority ancestry, you will have problems with accurate modelling, but I am sure G25 works perfect for completely unmixed people
    Last edited by CyrylBojarski; 04-21-2021 at 04:12 PM.

  19. The Following User Says Thank You to CyrylBojarski For This Useful Post:

     altvred (04-21-2021)

Page 1 of 2 12 LastLast

Similar Threads

  1. Single thread for G25 modelling requests
    By crazyladybutterfly in forum Autosomal (auDNA)
    Replies: 224
    Last Post: Yesterday, 04:43 PM
  2. Experiment with Vahaduo modern scaled Distance Feature
    By MitchellSince1893 in forum Autosomal (auDNA)
    Replies: 44
    Last Post: 04-22-2021, 10:18 PM
  3. Replies: 9
    Last Post: 05-18-2017, 10:34 PM
  4. Replies: 0
    Last Post: 09-22-2016, 11:05 AM
  5. Genetic distance vs geographic distance
    By Afshar in forum Other
    Replies: 0
    Last Post: 07-17-2016, 02:19 PM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •