Announcement

Collapse
No announcement yet.

Population Finder - what to choose?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • vinnie
    replied
    Originally posted by vinnie View Post
    More new results tonight, this time for Mom.

    Original Illumina PF: 100% Middle Eastern Jewish MOE .01%

    tonight:

    Middle East Bedouin, Bedouin South, Druze, Jewish, Mozabite, Palestinian 59.48% MOE 22.44%

    Europe Sardinian, Spanish, Tuscan, French 40.52% MOE 22.44%

    With that high of a MOE, I expect further refinement. But these results seem more in keeping with Sicily's history. Additionally, I found her maiden name listed as a medieval Jewish French/Spanish surname in an old French surname book, so again, the two sources might be confirming each other.
    Forgot to add that one of her ancestors has a surname that is most likely of Spanish origin, and another that might be Arabic, but I've yet to research that. Possibly more cooperation between PF and the paper trail.

    Leave a comment:


  • Yaffa
    replied
    Originally posted by DelawareNative View Post
    I wish FTDNA would send me a heritage survey - the more we share, the more info they can sort out...then they can get this PF calibrated, and therefore we'll have accurate results - or am I oversimplifying it?
    Giving them your known ancestry will probably help but could also hinder. When I say this, it's not directed at you but a general statement. I have seen many people make up things on their ancestors that are not true. My ancestors have been mutilated with false info because there are many people who dont know how to research. I don't think FTDNA has the time to check and make sure everyone's trees are correct. This is why in order for FTDNA to get AI DNA it has to come from tribal members who are documented as Indian. There are plenty of people out there who claim Indian but have no documentation to prove it. Those people would not make good DNA test subjects for these kinds of tests. Most of the tribes with documentation will not give up their DNA and I dont blame them

    Leave a comment:


  • aidan
    replied
    Here are my revised Population Finder results

    Continent (Subcontinent) - Europe (Western European)
    Population: Orcadian - 99.20%
    http://i52.tinypic.com/10ojmg1.jpg

    Leave a comment:


  • jloe
    replied
    Latest w/survey:
    http://img560.imageshack.us/img560/2...nder462011.png

    Continent (Subcontinent) Population Percentage Margin of Error
    Europe Basque, French, Orcadian 68.77%
    Middle East (Jewish) Jewish 31.23%

    Leave a comment:


  • vinnie
    replied
    More new results tonight, this time for Mom.

    Original Illumina PF: 100% Middle Eastern Jewish MOE .01%

    tonight:

    Middle East Bedouin, Bedouin South, Druze, Jewish, Mozabite, Palestinian 59.48% MOE 22.44%

    Europe Sardinian, Spanish, Tuscan, French 40.52% MOE 22.44%

    With that high of a MOE, I expect further refinement. But these results seem more in keeping with Sicily's history. Additionally, I found her maiden name listed as a medieval Jewish French/Spanish surname in an old French surname book, so again, the two sources might be confirming each other.

    Leave a comment:


  • vinnie
    replied
    I've been surveyed for the second time in one week with new PF results tonight for my dad.

    last week - same as original Illumina PF results:
    Europe (Southern European) Tuscan 63.75% MOE 16.31%
    Middle East Druze, Jewish, Palestinian 36.25% MOE 16.31%

    tonight:
    Europe Basque, Spanish, Tuscan 58.19% MOE 14.78%
    Middle East Druze, Jewish, Palestinian 41.81% MOE 14.78%

    I'm glad the margin of error is coming down, and the Basque/Spanish is particularly interesting. His maternal grandmother was born out of wedlock, and years ago I found that she had listed a Spanish sounding surname as her father's name on her SSI application. After checking the Italian white pages, I found that the surname was concentrated mainly in the Venice area (far from where we're from) which historically had a strong Spanish presence. It looks like PF and the paper trail might be confirming each other.

    (MOE = margin of error) Thank you, Nathanm!
    Last edited by vinnie; 6 April 2011, 10:49 PM.

    Leave a comment:


  • nathanm
    replied
    Originally posted by vinnie View Post
    One more try...Nope, didn't work.
    Certain characters screw up the software running this forum. I tried copy/paste directly from my PF results, but the single character for +/- would cut off the rest of the post. Lots of people have had similar experiences.

    Leave a comment:


  • vinnie
    replied
    Originally posted by vinnie View Post
    Help! Can't post! What's happening?
    One more try...Nope, didn't work.

    Leave a comment:


  • vinnie
    replied
    Help! Can't post! What's happening?

    Leave a comment:


  • vinnie
    replied
    something weird's happening to Forum posting...
    Last edited by vinnie; 6 April 2011, 10:33 PM. Reason: something weird happening to the Forum posting system.

    Leave a comment:


  • nathanm
    replied
    Originally posted by DelawareNative View Post
    Yep, read his article on that. So how will the Population Finder become more accurate? What kind of data will it take to change things? Or, is it the interpretation of the data?
    I think interpretation is part of the answer. The way FF works, and the differences between it, 23andMe's Ancestry Painting and Global Similarity (advanced view), and third party tools like Dr McDonald's still isn't clear to me. I've read the entire FAQ, and follow conversations here and on dna-forums, but still don't think I understand well enough. Is it an average, or do we have SNPs in common with those populations? It's not being explained well.

    The data's already out there to make it better. FTDNA's still using the HGDP for reference populations, which was probably the best available when they first created PF. But now the HapMap and 1000 Genomes projects have more comprehensive reference samples. I think they should either switch to using one of them, or create a composite of all of the above.

    Leave a comment:


  • vinnie
    replied
    message cut off
    Last edited by vinnie; 6 April 2011, 10:34 PM. Reason: message was cut off in the posting

    Leave a comment:


  • DelawareNative
    replied
    Yep, read his article on that. So how will the Population Finder become more accurate? What kind of data will it take to change things? Or, is it the interpretation of the data?

    Leave a comment:


  • nathanm
    replied
    Originally posted by DelawareNative View Post
    I wish FTDNA would send me a heritage survey - the more we share, the more info they can sort out...then they can get this PF calibrated, and therefore we'll have accurate results - or am I oversimplifying it?
    It's not that simple. Especially for typical Americans who are mutts like me. The best reference populations have stayed in the same area for several generations, and generally marry within the same ethnic group. It's far more difficult to analyze admixed populations. Dienekes had an interesting blog post on this using Mexico and India as examples.

    Leave a comment:


  • DelawareNative
    replied
    I wish FTDNA would send me a heritage survey - the more we share, the more info they can sort out...then they can get this PF calibrated, and therefore we'll have accurate results - or am I oversimplifying it?

    Leave a comment:

Working...
X