Page 1 of 17 12311 ... LastLast
Results 1 to 10 of 167

Thread: From Mauri Myllylä and Joe Pickrell: an ancestry calculator for Linux users

  1. #1
    Gold Class Member
    Posts
    7,572
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA (P)
    R-BY3604-Z275
    mtDNA (M)
    H5a1

    Normandie Orkney Netherlands Friesland East Frisia Finland

    From Mauri Myllylä and Joe Pickrell: an ancestry calculator for Linux users

    Thanks to Helgenes50, I've recently discovered an ancestry calculator that I had missed on Mauri Myllylä's blog: http://terheninenmaa.blogspot.fr/201...s-revised.html
    Despite of the title of Mauri's post, it's not strictly speaking an "admixture calculator", for the simple reason that it does'nt use ADMIXTURE, but Joe Pikrell's algorithm. It may primarily interest north-european people, as the southern populations are not at all detailed (one only group "mediterranean"). Another limitation, perhaps critical: it works only on Linux ( Python necessary, and a complete gsl library). Each analysis takes more than 4 hours on my computer (but likely less on more recent devices, mine is 6 years old). The procedure is in principle very easy, worked immediatly on my system, but so far refuses to work on Helgenes' one, and it's one of the reasons why I open the thread: how will it work on other systems? I've tried it on my parents' genomes and mine (phased). Here are the results (without details, each result is the average of 10 bootstrap outputs). Overlapping on ~500.000 snps.

    Father:
    Northwest_European 25.8
    East_Scandinavian 19.4
    Baltic 18.4
    Central_European 16.8
    Mediterranean 12.5
    Saami 5.3
    Finnic 1.8

    Mother:
    East_Scandinavian 37.3
    Central_European 29.2
    Northwest_European 25.1
    Slavic 5.2
    Mediterranean 2.0
    AMBIG_East_European 1.1

    Me:
    East_Scandinavian 28.3
    Central_European 25.2
    Northwest_European 22.7
    Mediterranean 7.9
    Baltic 7.9
    Slavic 3.5
    Saami 2.4
    Finnic 1.3

    I hope that at least 2 or 3 among you friends will try this.
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  2. The Following 13 Users Say Thank You to anglesqueville For This Useful Post:

     A Norfolk L-M20 (01-24-2018),  Amerijoe (01-19-2018),  Camulogène Rix (01-20-2018),  cercle (01-19-2018),  evon (01-19-2018),  Helgenes50 (01-19-2018),  karwiso (01-20-2018),  khanabadoshi (01-24-2018),  Massam (01-19-2018),  michal3141 (01-19-2018),  randwulf (01-19-2018),  Rigel (01-19-2018),  sktibo (01-20-2018)

  3. #2
    Registered Users
    Posts
    2,377
    Sex
    Location
    Western Norway
    Ethnicity
    Norwegian
    Nationality
    Norwegian
    Y-DNA (P)
    R1b-FT71931
    mtDNA (M)
    U5b1b1(T16192C)
    Y-DNA (M)
    R1b-U152
    mtDNA (P)
    H13a1a

    Quote Originally Posted by anglesqueville View Post
    Thanks to Helgenes50, I've recently discovered an ancestry calculator that I had missed on Mauri Myllylä's blog: http://terheninenmaa.blogspot.fr/201...s-revised.html
    Despite of the title of Mauri's post, it's not strictly speaking an "admixture calculator", for the simple reason that it does'nt use ADMIXTURE, but Joe Pikrell's algorithm. It may primarily interest north-european people, as the southern populations are not at all detailed (one only group "mediterranean"). Another limitation, perhaps critical: it works only on Linux ( Python necessary, and a complete gsl library). Each analysis takes more than 4 hours on my computer (but likely less on more recent devices, mine is 6 years old). The procedure is in principle very easy, worked immediatly on my system, but so far refuses to work on Helgenes' one, and it's one of the reasons why I open the thread: how will it work on other systems? I've tried it on my parents' genomes and mine (phased). Here are the results (without details, each result is the average of 10 bootstrap outputs). Overlapping on ~500.000 snps.

    Father:
    Northwest_European 25.8
    East_Scandinavian 19.4
    Baltic 18.4
    Central_European 16.8
    Mediterranean 12.5
    Saami 5.3
    Finnic 1.8

    Mother:
    East_Scandinavian 37.3
    Central_European 29.2
    Northwest_European 25.1
    Slavic 5.2
    Mediterranean 2.0
    AMBIG_East_European 1.1

    Me:
    East_Scandinavian 28.3
    Central_European 25.2
    Northwest_European 22.7
    Mediterranean 7.9
    Baltic 7.9
    Slavic 3.5
    Saami 2.4
    Finnic 1.3

    I hope that at least 2 or 3 among you friends will try this.
    I have never run this sort of program, could you give me a step by step guide? Ive installed Python3.6 and have the files ready..what do I do next?

  4. #3
    J Man
    Guest
    Seems interesting.

  5. #4
    Registered Users
    Posts
    939
    Sex
    Location
    Krakow
    Ethnicity
    Polish
    Nationality
    Pole
    Y-DNA (P)
    R1a > M458 > Y23110
    mtDNA (M)
    J1c1

    Poland
    Quote Originally Posted by evon View Post
    I have never run this sort of program, could you give me a step by step guide? Ive installed Python3.6 and have the files ready..what do I do next?
    Instructions are available:

    Program downloading and running

    Download programs here. Unzip and locate all programs into a same directory. To run tests you need use a command line "bash ./ajo1.sh <sample-id>, where sample-id is the file name holding your genetic data in 23andme format. The sample file must be compressed with gz file extension (gzip format), but on the command line you give only the sample id (sample-id.gz), not the extension. The test works fine with following genome builds: HG18, HG19, GRCh36, GRCh37, but if your genome file is in the FtDna format you have to convert it into the 23andme style. On Linux it is done easily using four command line entries:

    first unzip your genome file and then

    cp <original filename> <sample-id>
    sed -i 's/"//g' <sample-id>
    sed -i 's/,/\t/g' <sample-id>
    gzip <sample-id>

    If your data is already in the 23andme format, but not compressed with gz file extension then you need to unzip it first and run the first and fourth commands explained as above.

  6. The Following User Says Thank You to michal3141 For This Useful Post:

     evon (01-19-2018)

  7. #5
    Registered Users
    Posts
    939
    Sex
    Location
    Krakow
    Ethnicity
    Polish
    Nationality
    Pole
    Y-DNA (P)
    R1a > M458 > Y23110
    mtDNA (M)
    J1c1

    Poland
    My results:

    Slavic 44.7
    Baltic 27.5
    East_Scandinavian 8.5
    Northeast_European 5.8
    Mediterranean 5.2
    Central_European 3.1
    Uralic 2.3
    Northwest_European 1.7
    AMBIG_European 1.2

    I think it is a very cool calculator.
    Last edited by michal3141; 01-19-2018 at 02:17 PM.

  8. The Following 3 Users Say Thank You to michal3141 For This Useful Post:

     Amerijoe (01-19-2018),  cercle (01-19-2018),  Helgenes50 (01-19-2018)

  9. #6
    Gold Class Member
    Posts
    7,572
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA (P)
    R-BY3604-Z275
    mtDNA (M)
    H5a1

    Normandie Orkney Netherlands Friesland East Frisia Finland
    Evon, take care, there must be NO QUOTE BEFORE THE COMMAND LINE:
    bash ./ajo1.sh <sample-id>
    All in one, starting from a file pat.txt (23&me format), in the folder where are Mauri's files ( admixtext2.gz, ancestry,grouping,ajo1.sh, summarize_bootstrap.py)
    You have to get the authorisation of running as a program (don't know how they tell that in english, or you in norwegian,, sure you know what I mean) for ajo1.sh.. Then, on a terminal:

    cp pat.txt pat
    gzip pat
    bash ./ajo1.sh pat ( a space between "bash" and "ajo1)

    If you use a ftdna file you can also convert it with aconv (from Felix), it works, I tried it. You get in your folder the results of each bootstrap, but, take care, they are erased at the end of the run, therefore if you want to keep them you have to do it before.
    Michal, how long dit it take for you?
    Last edited by anglesqueville; 01-19-2018 at 02:33 PM.
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  10. The Following User Says Thank You to anglesqueville For This Useful Post:

     evon (01-19-2018)

  11. #7
    Gold Class Member
    Posts
    7,572
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA (P)
    R-BY3604-Z275
    mtDNA (M)
    H5a1

    Normandie Orkney Netherlands Friesland East Frisia Finland
    So far the only results I can speak of, are those of my parents, and me. It looks impressively coherent. For my father, the ancestry "saami" does appear at more or less high levels with many calculators, and I verified some months ago its probable reality with D-stats, Pagani, and my dad's imputed datas, on around 5 millions snps. What is curious here is that his "saami" score is higher than his "finnic", perhaps an issue with the east-scandinavian datas ( north-swedes from the Behar study). I agree with Michal, it's likely a very interesting tool, even if its "audience" is fatally limited.
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  12. The Following User Says Thank You to anglesqueville For This Useful Post:

     Helgenes50 (01-19-2018)

  13. #8
    Registered Users
    Posts
    939
    Sex
    Location
    Krakow
    Ethnicity
    Polish
    Nationality
    Pole
    Y-DNA (P)
    R1a > M458 > Y23110
    mtDNA (M)
    J1c1

    Poland
    Quote Originally Posted by anglesqueville View Post
    Michal, how long dit it take for you?
    Less than 1h but don't know exactly. I didn't measure time. I was running this on my brand new intel i7 quad-core laptop. Also I was using AncestryDNA raw data which I converted into 23andMe format. I suspect that the running time may also depend on the original source of your raw data.

  14. #9
    Gold Class Member
    Posts
    7,572
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA (P)
    R-BY3604-Z275
    mtDNA (M)
    H5a1

    Normandie Orkney Netherlands Friesland East Frisia Finland
    Quote Originally Posted by michal3141 View Post
    Less than 1h but don't know exactly. I didn't measure time. I was running this on my brand new intel i7 quad-core laptop. Also I was using AncestryDNA raw data which I converted into 23andMe format. I suspect that the running time may also depend on the original source of your raw data.
    I have to buy a new computer, indeed. I have an old AMD Phenom II at 2.8 Ghz ( and furthermore my fedora is installed on a virtual machine, not very good for the computing speed...)
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  15. #10
    Registered Users
    Posts
    2,377
    Sex
    Location
    Western Norway
    Ethnicity
    Norwegian
    Nationality
    Norwegian
    Y-DNA (P)
    R1b-FT71931
    mtDNA (M)
    U5b1b1(T16192C)
    Y-DNA (M)
    R1b-U152
    mtDNA (P)
    H13a1a

    Quote Originally Posted by anglesqueville View Post
    I have to buy a new computer, indeed. I have an old AMD Phenom II at 2.8 Ghz ( and furthermore my fedora is installed on a virtual machine, not very good for the computing speed...)
    Mine takes about 10 mins per run, so that will be 10 x 10 in total I guess.. Running Ubuntu 16.04, on a AMD Athlon(tm) X4 860K Quad Core Processor × 4 with 16gb ram..
    Last edited by evon; 01-19-2018 at 03:11 PM.

  16. The Following User Says Thank You to evon For This Useful Post:

     anglesqueville (01-19-2018)

Page 1 of 17 12311 ... LastLast

Similar Threads

  1. BAM Analysis Kit (Windows and Linux). New release.
    By teepean47 in forum Open-Source Projects
    Replies: 70
    Last Post: 10-05-2020, 09:58 AM
  2. Replies: 82
    Last Post: 07-03-2018, 12:07 AM
  3. "From Mauri Myllylä and Joe Pickrell ... ", Season II: The Ancients
    By anglesqueville in forum Autosomal (auDNA)
    Replies: 28
    Last Post: 01-29-2018, 04:49 PM
  4. Replies: 14
    Last Post: 11-04-2016, 09:33 PM

Posting Permissions

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