Announcement

Collapse
No announcement yet.

Fun things to do with Family Finder raw data?

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

  • Fun things to do with Family Finder raw data?

    I'm thrilled to have received my Family Finder results three weeks early (batch 402). I have 12 close matches and over 100 speculative matches. So far, I've only been able to identify one match through the genealogical paper trail - a speculative one at the 4th cousin - distant level.

    I've discovered Promethease (http://www.snpedia.com/index.php/Promethease) and had quite a bit of fun with it. It was remarkably on-target with what I know about my family's medical history.

    I've also been to GedMatch.com (http://www.gedmatch.com) and discovered some new relations - a few estimated 3rd cousins.

    Can anyone recommend any other sources for analysis of our raw Family Finder autosomal data?

    Jennifer

  • #2
    dbSNP

    I like to look at my raw data, pick a snp or two an look them up in dbSNP, the National Institutes of Health database of everything snp-ish.


    You pick something like chromosome 9's rs10867105, and enter it into the search at the url below. You find out things - is your value typical of Europeans? Or sub-saharan Africa? How many samples have been filed on it? You may have to stretch to understand it. It may even tell you what that snp does. It also has the dna sequence surrounding the snp which is how the values are tallied anyway. SNP pedia that works with Promethease is a similar private database. I have found a few snp's where I have odd values for my generally understood background and one snp where my value had not been recorded.




    http://www.ncbi.nlm.nih.gov/projects/SNP/

    Comment


    • #3
      Well since FTDNA now uses the same chip as 23andme you can probably take advantage of these third party tools:

      http://www.23andyou.com/3rdparty

      Comment


      • #4
        Originally posted by jennyology View Post
        I'm thrilled to have received my Family Finder results three weeks early (batch 402). I have 12 close matches and over 100 speculative matches. So far, I've only been able to identify one match through the genealogical paper trail - a speculative one at the 4th cousin - distant level.

        I've discovered Promethease (http://www.snpedia.com/index.php/Promethease) and had quite a bit of fun with it. It was remarkably on-target with what I know about my family's medical history.

        I've also been to GedMatch.com (http://www.gedmatch.com) and discovered some new relations - a few estimated 3rd cousins.

        Can anyone recommend any other sources for analysis of our raw Family Finder autosomal data?

        Jennifer
        Aside from those, I'd also mention Dr McDonald.

        I downloaded Promethease the other day, and input my FF data (my father already got similar analysis from 23andMe). It's interesting, and SNPedia looks like it has a lot of potential, but there's just not enough content to be truly useful yet. Specifically, I was looking at the various genes for eye color (there are several interrelated). The explanations on SNPedia were slim at best, and often confusing. Also, I think the data from the various companies isn't reported in the same format. Several SNPs list the predominant values in different populations and overall, for example C:C, and a minority with C:T or T:T. But on that SNP, my data said G:G. Apparently that's my alien DNA.

        Comment


        • #5
          Wow, these look cool, but Enlis and Genomera only accept data in 23andme format. Is there a utility to convert between formats?
          Last edited by brookes; 20 April 2011, 12:08 AM.

          Comment


          • #6
            Originally posted by nathanm View Post
            Aside from those, I'd also mention Dr McDonald.

            I downloaded Promethease the other day, and input my FF data (my father already got similar analysis from 23andMe). It's interesting, and SNPedia looks like it has a lot of potential, but there's just not enough content to be truly useful yet. Specifically, I was looking at the various genes for eye color (there are several interrelated). The explanations on SNPedia were slim at best, and often confusing. Also, I think the data from the various companies isn't reported in the same format. Several SNPs list the predominant values in different populations and overall, for example C:C, and a minority with C:T or T:T. But on that SNP, my data said G:G. Apparently that's my alien DNA.
            Re: Data not in same format... see next.

            Comment


            • #7
              Originally posted by brookes View Post
              Wow, these look cool, but Enlis and Genomera only accept data in 23andme format. Is there a utility to convert between formats?
              (edited) nevermind.. I tried it and it failed for me too.
              Last edited by mkdexter; 20 April 2011, 01:09 AM.

              Comment


              • #8
                Originally posted by mkdexter View Post
                Try your raw data download again. Sometime after first allowing downloads, FTDNA went back and reformatted so their raw data format matches 23andMe. At least it appears that way. It changed and I had to re-download mine because when I tried to compare the original download with a family member's new test it failed. After I re-downloaded mine my comparison worked agin.
                O rly.. Well I just went and tried again and I still get "Error - import failed". I'm using a fresh copy of my raw data, unzipped. The Quick Start for Enlis personal says it's expecting a txt file, should I rename my .csv to txt and try again? Has anyone done this in Enlis Personal?

                Comment


                • #9
                  Yes I had also tried it and it failed so I edited my post.. I think its in the header of the file where there is a problem. It's looking for a certain pre-data header to be there.

                  Comment


                  • #10
                    My theory was wrong. So it is the build after all..

                    I had heard that FTDNA went back and redid their format so that it call came out dbSNP plus orientation. (I actually have not looked, ) The old way was plus or minus. Maybe this one Enlis program needs work.

                    I think FTDNA did reformat the raw data because my mother's test and mine compared to some new tests did not work until I re-downloaded my raw data again.

                    MD.
                    Last edited by mkdexter; 20 April 2011, 01:31 AM.

                    Comment


                    • #11
                      Originally posted by mkdexter View Post
                      My theory was wrong. So it is the build after all..

                      I had heard that FTDNA went back and redid their format so that it call came out dbSNP plus orientation. (I actually have not looked, ) The old way was plus or minus. Maybe this one Enlis program needs work.

                      I think FTDNA did reformat the raw data because my mother's test and mine compared to some new tests did not work until I re-downloaded my raw data again.

                      MD.
                      FTDNA has fixed the autosomal files, but not the X-chromosome raw data file. GEDmatch applies a dbSNP correction to the X-chromosome data when you upload it. If you upload your Illumina X-chromosome raw data to a site like Prometheus, it will not give correct results.

                      Comment


                      • #12
                        I've uploaded my raw data into gedmatch.com and I went from having zero match on here to having a bunch over there. I'm a bit overwhelmed how accurate are these matches. Is there a way for me to tell how far back it goes or possible relationships we may have (aside from e-mailing everyone individually and asking them)?

                        Comment


                        • #13
                          Never mind. I've figured it out.

                          Originally posted by ahernandez View Post
                          I've uploaded my raw data into gedmatch.com and I went from having zero match on here to having a bunch over there. I'm a bit overwhelmed how accurate are these matches. Is there a way for me to tell how far back it goes or possible relationships we may have (aside from e-mailing everyone individually and asking them)?

                          Comment


                          • #14
                            Originally posted by ahernandez View Post
                            I've uploaded my raw data into gedmatch.com and I went from having zero match on here to having a bunch over there. I'm a bit overwhelmed how accurate are these matches. Is there a way for me to tell how far back it goes or possible relationships we may have (aside from e-mailing everyone individually and asking them)?
                            Most of the comparison results at GEDmatch do NOT qualify as 'matches'. GEDmatch reports results far below the thresholds set by FTDNA and 23andMe. An easy rule-of-thumb is to treat anything with a blank in the 'gen' column as a non-match. The 'gen' number is an estimate of the number of generations back to the common ancestor. However, it loses a LOT of accuracy past about the 4th generation, and that should be part of your consideration when using it.

                            I DEFINITELY DO NOT recommend emailing everybody and asking them if they are related to you. THAT IS SPAM. DON'T DO IT! Please be responsible and limit your emails to your CLOSE relatives.

                            Comment


                            • #15
                              Got it. I wasn't planning on e-mailing everyone. I did e-mail the two two people with the MRCA at 5 gen, which now after reading your post I see was probably a waste of time. Thanks for the heads up.

                              Comment

                              Working...
                              X