Announcement

Collapse
No announcement yet.

Experiences with Family Finder and PF

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

  • Experiences with Family Finder and PF

    I would be interested in knowing how Family Finder and other DNA ancestry tests are helping other people in investigating their ancestry. Here are some of my experiences:

    Here is a comparison between my PF and an autosomal DNA fingerprint test I did about 2 yrs ago. There are definitely some similarities between them indicating the "Heinz 57 varieties" in my ancestry. My known ancestry, from a fairly extensive family tree: Colonial Quebec, Colonial New England, Colonial Southern (AL, TN, GA), English (1870's) and Cherokee. The main piece of information in PF that is missing is the Native American. I was hoping PF would give me a more accurate percentage of Native American than my crude estimate of 1/8, but instead it completely overlooks it. I would say that the PF takes a very shallow, blurry look at ancestry and the fingerprint takes a deep, sharp focus but is not very thorough. The fingerprint test I found very useful in providing clues to my tree. For example, it found gypsy heritage that I didn't know of, and on investigating I found out one of my great aunts was a "famous" fortune teller and psychic in the 1890's in Boston and San Francisco. The PF results are not very useful so far except in giving an overall impression of my heritage, but it seems to fit well with the other test in a general way.

    I'm getting a lot of matches (two 3rd and many 4th cousins). All of the 7 people I've communicated with have responded. So far I have not been able to identify any definite connections to any of them, although there are some hints that surnames and localities are similar. I'm suspecting there are a some disconnects between legal and biological parents in my tree. I did notice that one of my GGGrandmothers in the south had a child 6 months after marrying, and that husband left her at that time. She later married another man, and I seem to have matches with his surname of Cross, but his ancestry is unknown. I hoping to sort this out when more matches come in. So far I'm happy with the matches and it should help given some time for FF to mature.
    Last edited by desley; 24 March 2011, 01:07 PM. Reason: original too long

  • #2
    addendum

    My results were cut off, here they are:

    FTDNA Population Finder:
    Europe (Western European) French, Orcadian, Spanish 84.74%
    Europe Tuscan, Finnish, Romanian, Russian 15.26%

    DNA Fingerprint Ethnotyping (not from FTDNA)
    Rank 13-Marker Top Matches out of 380+
    Populations (with Key to Data) RMP
    1 Portuguese (6) 4.49E+15
    2 Byelorussian (163) 1.08E+16
    3 Michigan Nat.Am. (2) 1.14E+16
    4 Polish (176) 1.19E+16
    5 Polish (217) 1.63E+16
    6 Tunisian (59) 1.81E+16
    7 Serbian (157) 1.90E+16
    8 Bosnian (156) 1.93E+16
    9 British (298) 2.09E+16
    10 California Caucasian (2) 2.18E+16
    11 Slovenian (222) 2.21E+16
    12 German (296) 2.26E+16
    13 Polish (162) 2.46E+16
    14 Azores (82) 2.57E+16
    15 Polish (Old Believers) (201) 2.58E+16
    16 Spanish (103) 3.10E+16
    17 Hungarian (Ashkenazim) (111) 3.22E+16
    18 Hungarian (Eastern Romanies) (111) 3.40E+16
    19 Bhumihar Brahmin (India) (72) 3.53E+16
    20 Southern Croatian (54) 3.81E+16
    Last edited by desley; 24 March 2011, 01:12 PM.

    Comment


    • #3
      I'm having the same problem - am Native, but not even a perceptible amount is coming up. Something is amiss, and other threads have addressed this...although FTDNA has not.

      Comment


      • #4
        Fingerprint test vs FTdna

        Originally posted by desley View Post
        I would be interested in knowing how Family Finder and other DNA ancestry tests are helping other people in investigating their ancestry. Here are some of my experiences:

        Here is a comparison between my PF and an autosomal DNA fingerprint test I did about 2 yrs ago. There are definitely some similarities between them indicating the "Heinz 57 varieties" in my ancestry. My known ancestry, from a fairly extensive family tree: Colonial Quebec, Colonial New England, Colonial Southern (AL, TN, GA), English (1870's) and Cherokee. The main piece of information in PF that is missing is the Native American. I was hoping PF would give me a more accurate percentage of Native American than my crude estimate of 1/8, but instead it completely overlooks it. I would say that the PF takes a very shallow, blurry look at ancestry and the fingerprint takes a deep, sharp focus but is not very thorough. The fingerprint test I found very useful in providing clues to my tree. For example, it found gypsy heritage that I didn't know of, and on investigating I found out one of my great aunts was a "famous" fortune teller and psychic in the 1890's in Boston and San Francisco. The PF results are not very useful so far except in giving an overall impression of my heritage, but it seems to fit well with the other test in a general way.

        I'm getting a lot of matches (two 3rd and many 4th cousins). All of the 7 people I've communicated with have responded. So far I have not been able to identify any definite connections to any of them, although there are some hints that surnames and localities are similar. I'm suspecting there are a some disconnects between legal and biological parents in my tree. I did notice that one of my GGGrandmothers in the south had a child 6 months after marrying, and that husband left her at that time. She later married another man, and I seem to have matches with his surname of Cross, but his ancestry is unknown. I hoping to sort this out when more matches come in. So far I'm happy with the matches and it should help given some time for FF to mature.
        I did the fingerprint test too and showed native american.
        With me I assumed that result was quite skewed - it is quite possible I have NA via the american whalers living in the Azores where my jewish Spanish ancestors went during the inquisition. But surely that would have been just a little bit or a complete long shot.
        It's probably also where my Belgium and Morrocan (Moorish prisoners) results and sub-suharan markers come from. They left there about 1800 so highly skewed? .... maybe ...or maybe not.
        The fingerprint showed me what my bits and pieces are and also showed my dominant ancestry (scottish)
        Only looked at the top 10 results and ignored the rest.
        I was very confused when I got those results ....but they make sense as I've made progress with the paper trail.

        However the NA has always puzzled me. The real benefit of Ftdna is that I'm getting some NA 4th cousin matches which hopefully will allow me to get to the bottom of this

        Comment


        • #5
          confirmation

          I can't speak to Family Finder since I've yet to get any matches but Population Finder pretty much confirmed my family's paper trail.

          I got 78% Western European (Spanish, French) and 21% Middle Eastern (Mozabite), which is exactly would a Canary Islander would get.

          Comment


          • #6
            Originally posted by ahernandez View Post
            I can't speak to Family Finder since I've yet to get any matches but Population Finder pretty much confirmed my family's paper trail.

            I got 78% Western European (Spanish, French) and 21% Middle Eastern (Mozabite), which is exactly would a Canary Islander would get.
            Have you tested with 23andMe? If so, how many and what kinds of matches have you seen with people in their database?

            Comment


            • #7
              Originally posted by jah View Post
              Have you tested with 23andMe? If so, how many and what kinds of matches have you seen with people in their database?
              No, I haven't. Family Finder is the first test of its kind that I've taken.

              Comment


              • #8
                I would not put too much weight to Population Finder at this time. Just my opinion, but I think it needs work to get to the level of being useful. Regarding your Native American ancestry, I don't think PF is good at sussing out smaller (<10%) components of ancestry. For example, I am mostly European ancestry but have 0.5% to 1% Sub Saharan African (SSA) ancestry in a few distinct segments, the largest of which is around 16 mega base pairs in length. My aunt is 1% to 2% SSA with the largest segment around 40 mega base pairs in length. PF showed no hint of this, but every other tool has (23andMe Ancestry Painting, Eurogenes project, and Dr. McDonald's tools). Heck, for that matter, one of my grandparents was born in Latvia (ethnic Lett) and PF didn't even show a hint at my 25% Baltic ancestry.

                I recommend you send your raw data to Dr. McDonald. His chromosome painter and 'spot on the map' analysis are very good.

                Family Finder on the other hand is very useful. I have found several cousins where we were able to identify the common ancestors, and have gotten clues about where the match may be in others where we couldn't find the common ancestors.

                Comment


                • #9
                  Originally posted by smallaxe View Post
                  I would not put too much weight to Population Finder at this time. Just my opinion, but I think it needs work to get to the level of being useful. Regarding your Native American ancestry, I don't think PF is good at sussing out smaller (<10%) components of ancestry. For example, I am mostly European ancestry but have 0.5% to 1% Sub Saharan African (SSA) ancestry in a few distinct segments, the largest of which is around 16 mega base pairs in length. My aunt is 1% to 2% SSA with the largest segment around 40 mega base pairs in length. PF showed no hint of this, but every other tool has (23andMe Ancestry Painting, Eurogenes project, and Dr. McDonald's tools). Heck, for that matter, one of my grandparents was born in Latvia (ethnic Lett) and PF didn't even show a hint at my 25% Baltic ancestry.

                  I recommend you send your raw data to Dr. McDonald. His chromosome painter and 'spot on the map' analysis are very good.

                  Family Finder on the other hand is very useful. I have found several cousins where we were able to identify the common ancestors, and have gotten clues about where the match may be in others where we couldn't find the common ancestors.
                  Thanks! Just took your advice and sent my results to Dr. McDonald. (Sure hope I sent the file in the proper format

                  Comment


                  • #10
                    I've sent my raw data to Dr. McDonald....and while I don't question his analysis, I question the integrity of my raw data - no perceptible NA??? I've sent feedback to FTDNA

                    Comment


                    • #11
                      Native American surprises

                      Originally posted by DelawareNative View Post
                      I've sent my raw data to Dr. McDonald....and while I don't question his analysis, I question the integrity of my raw data - no perceptible NA??? I've sent feedback to FTDNA
                      Strange, DelawareNative. It sounds like you expected some NA ancestry from Dr. McDonald's data and got none.

                      I, on the other hand, was shocked to get .03 %. He describes it as "small but strong" and believes is probably reflects something real.

                      The possibility is fascinating but I struggle to explain it, with my ancestors all having arrived relatively recently from Europe 1898-1920.

                      Just one odd thing, which I thought was a mistake: A reference to Finnish on my grandmother's birth certificate! (Finnish/Sammi can show up as N.A. apparently.)

                      Have to think about this!

                      Comment


                      • #12
                        Originally posted by bkilpatrick View Post
                        I, on the other hand, was shocked to get .03 %. He describes it as "small but strong" and believes is probably reflects something real.
                        Correction: I meant to write .3% NA.

                        Comment


                        • #13
                          Where do you find out about sending the raw files to Dr McDonald?

                          Comment


                          • #14
                            Same question and...

                            "Where do you find out about sending the raw files to Dr McDonald?

                            Yes, and what is the process of sending raw data to Dr. McDonald and are we talking about FF raw data, Mtdna or Ytdna results? If they are included, do they have to be tested at the FGS level?

                            My PF results are not accurate by any means, and am wondering if the FF raw data can be better analyzed. I hope so.

                            Comment


                            • #15
                              Originally posted by Darkwriter View Post
                              [COLOR="Blue"]My PF results are not accurate by any means, and am wondering if the FF raw data can be better analyzed. I hope so.
                              For non-mixed people, there are ways to estaminate their origin quiet well. If their anchestors havnt left the local region for 6 or so generations, possibly even to a spot of 100km diameter.

                              But once a person is mixed, its a lot harder to estaminate the origins.

                              Thats the problem of ALL the projects out there and all known methods and I would say, a satisfying solution is not found yet.

                              The new version of the software "admixture", Version 1.1 has a new function for this (Estaminate Admixture level in individuals who are known to be admixed and whos admixture components are known.)

                              But even on 50/50 and using exactly those populations as references, that took place in this persons admixture (means: "Ideal conditions"), it returns results like 40/60


                              And now imagine the accuracity of cases in wich 3, 4 or 5 countries are intermixed and where half of these are not known or wich arent part of the reference populations.
                              Last edited by Daniel72; 26 March 2011, 04:27 AM.

                              Comment

                              Working...
                              X