Hi Terry, I like your approach, but I think that using countries as a "unit of reference" is misleading. Some countries are big, other are small; some have lots of (extented) DNA-samples others few ... The problem to overcome the bias created by the succes of DNA testing in the anglosaxon word and the fear of ADN tests in France is not an easy one. In case of I2.0110* (Hg I-L39 aka I-L38) there are indications that its resolution in the Alsace (bordering Germany) and Northern France (bordering Flanders) are higher. I have no indications that its frequency in the other French regions has a similar (higher) frequency. Tracing I-L38 family lines mostly results in a direct or indirect link to the Rhine region. Looking at the frenchheritage FTDNA project group I only can conclude that the % of I-L39 (the projects still calls them I2b2) in France (as a whole) is low. http://www.familytreedna.com/public/frenchheritage/default.aspx?vgroup=frenchheritage&vgroup=frenchheritage&vgroup=frenchheritage&vgroup=frenchheritage&section=yresults Best, Hans > Date: Fri, 24 Feb 2012 16:12:24 +1100 > From: tdrobb@gmail.com > To: y-dna-haplogroup-i@rootsweb.com > Subject: Re: [yDNAhgI] STR Branches, SNP's, Geographic Origin > > Daniel, > The data sources I used, for the STR 67-markers, are from various public > FTDNA Projects (mainly the I1 Project), plus data from Ysearch (making sure > not to double count the same data). > > The maps are the *frequency* of occurrence of the particular branch within > the given country/region. Here the frequency is the number of people in the > branch, divided by the number of people that have been tested in the > region. That weighting hopefully reduces the bias of popular DNA testing > regions like England, compared to some less popular testing regions. > > For your I2.0110* Branch, that frequency comes out highest in France and > Switzerland. Numerically, there are a large number of I2.0110* people with > a reported origin from England, but that large number should be scaled down > by the correspondingly large number of people from England that have taken > the STR 67-marker test. After that is accounted for, places like France and > Switzerland come out the dominant (highest frequency) place. But England, > and even Norway, still have a significant frequency of I2.0110*. (In > contrast, the I2.00* branch for example, has a very low frequency in > England, and a high frequency in Eastern Europe.) > > Terry > > 2012/2/23 Daniel Suárez Díaz <dsuarez_job@hotmail.com> > > > > > Terry, > > > > Thank you for share your work. Let me make you some questions about the > > branch I2.0110*. These branch correspon to the I2ab2 haplogroup and > > acording to the studies of Hans de Beule, the present day distribution is > > limited to the Upper Rhine and the British Isles. Also in the I2ab2 FDNA > > project, the most of the samples are British. However in your I2.0110* map > > with 63 samples, the major figures correspon to France and Switzerland. > > > > Therefore my questions are: > > > > - How many French and Swiss samples have you used? > > > > - Which are your sources? > > > > - And if the sources are public, Have you overweighted them or in other > > words, Have you underweighted the British samples because many of them are > > the same family? > > > > Thanks in advance. > > > > Daniel Suarez. > > > > > > > > > > > Date: Thu, 23 Feb 2012 01:27:04 +1100 > > > From: tdrobb@gmail.com > > > To: Y-DNA-HAPLOGROUP-I@rootsweb.com > > > Subject: [yDNAhgI] STR Branches, SNP's, Geographic Origin > > > > > > Often people ask about what SNP's they should look at testing, given > > their > > > particular STR values. In addition, people sometimes ask what can be said > > > about their ancestral geographic origin based again on their STR values. > > > > > > One can in fact answer those types of questions in an objective sense. > > > > > > The methodology I used is quite simple. First, I collected as much STR > > > haplotype data as I could (I got over eight thousand I1 and I2 samples > > from > > > various FTDNA Projects and from the Ysearch database). Next, I organised > > > that large STR dataset into "hierarchical clusters" using a standard > > > mathematical technique. Finally, I counted up the number of SNP alleles > > > (either positive or negative) that occurred within each cluster, now > > called > > > an STR "Branch", and I displayed the results; and I also plotted maps of > > > the frequency distribution of the most-distant-known-ancestor of each STR > > > Branch member across all regions of Europe. > > > > > > The results are very informative. > > > > > > The answer to many questions just depend on knowing your STR "Branch > > Code". > > > In the system I use, the Branch Code is essentially just a way of > > labelling > > > the STR branches in the computed hierarchical tree, and it is very > > similar > > > to the "Henry System" used in genealogy for numbering the descendants of > > an > > > ancestor. > > > > > > To determine your very own STR Branch Code in the system I use, see: > > > http://www.goggo.com/cgi-bin/branchFind.cgi > > > > > > Enter you FTDNA Kit Number or Ysearch ID, and (hopefully) it will return > > > your STR Branch Code. (Only works for y-haplogroups I1 and I2, and 67 or > > > more STR markers at present.) > > > > > > Follow the links to see what your STR Branch Code can tell you. > > > > > > Terry > > > > > > ------------------------------- > > > To unsubscribe from the list, please send an email to > > Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' > > without the quotes in the subject and the body of the message > > > > > > ------------------------------- > > To unsubscribe from the list, please send an email to > > Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' > > without the quotes in the subject and the body of the message > > > > ------------------------------- > To unsubscribe from the list, please send an email to Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' without the quotes in the subject and the body of the message

Hans, Thanks for your question, which does raise some important issues. First, I should say that I do have some smaller regional bins inside some of the countries (Germany, France, Spain, and Italy), but I am not currently plotting the data on that finer scale (yet) for a couple of reasons. If I were to plot the frequency of say I1 or I2 as a whole, than I have more than enough data with geographic origin, to make such maps. But when I split I1 and I2 up into the various branches, as determined by the STR tree, then the number of data points inside each regional bin can occasionally get a bit scarce, and so the corresponding frequency that is computed can get unstable. In the end, I ended up only doing regional bins for Scotland, England, and Wales. Next, the bias for high DNA testing countries/regions (such as England) is removed, because I always just do *frequency* maps. The frequency of occurrence of a particular DNA result, can be estimated for any country/region by dividing the raw count of people that have that particular result of interest, divided by the number of people that were tested - so the country population size, either large or small, is irrelevant and should not effect the frequency calculation. Finally, one should note that I am just presenting the results based on (hopefully) a clearly explained methodology. The method is simple, and the results are what they are. I am aware that the self-reported place of geographic origin, that I am obtaining from public sources, may not be totally reliable. That and other issues need to be remembered when reading the maps. But as I say, the results are what they are. With finer geographic binning, I would suspect you would be correct in saying that the I2.0110* branch would have a higher frequency in certain parts of France compared to others. I may work on that later, to see if there is enough data to make a better map. In the meantime, the I2.0110* map erroneously indicates that the frequency is uniform all over France because that is the binning unit I used there unfortunately. With time I will improve the maps. Terry On Mon, Feb 27, 2012 at 8:20 AM, Hans De Beule <hans.debeule@hotmail.com>wrote: > > Hi Terry, > > I like your approach, but I think that using countries as a "unit of > reference" is misleading. Some countries are big, other are small; some > have lots of (extented) DNA-samples others few ... The problem to overcome > the bias created by the succes of DNA testing in the anglosaxon word and > the fear of ADN tests in France is not an easy one. > > In case of I2.0110* (Hg I-L39 aka I-L38) there are indications that its > resolution in the Alsace (bordering Germany) and Northern France (bordering > Flanders) are higher. I have no indications that its frequency in the other > French regions has a similar (higher) frequency. Tracing I-L38 family lines > mostly results in a direct or indirect link to the Rhine region. > > Looking at the frenchheritage FTDNA project group I only can conclude that > the % of I-L39 (the projects still calls them I2b2) in France (as a whole) > is low. > > > http://www.familytreedna.com/public/frenchheritage/default.aspx?vgroup=frenchheritage&vgroup=frenchheritage&vgroup=frenchheritage&vgroup=frenchheritage&section=yresults > > Best, > > Hans > > > > Date: Fri, 24 Feb 2012 16:12:24 +1100 > > From: tdrobb@gmail.com > > To: y-dna-haplogroup-i@rootsweb.com > > Subject: Re: [yDNAhgI] STR Branches, SNP's, Geographic Origin > > > > Daniel, > > The data sources I used, for the STR 67-markers, are from various public > > FTDNA Projects (mainly the I1 Project), plus data from Ysearch (making > sure > > not to double count the same data). > > > > The maps are the *frequency* of occurrence of the particular branch > within > > the given country/region. Here the frequency is the number of people in > the > > branch, divided by the number of people that have been tested in the > > region. That weighting hopefully reduces the bias of popular DNA testing > > regions like England, compared to some less popular testing regions. > > > > For your I2.0110* Branch, that frequency comes out highest in France and > > Switzerland. Numerically, there are a large number of I2.0110* people > with > > a reported origin from England, but that large number should be scaled > down > > by the correspondingly large number of people from England that have > taken > > the STR 67-marker test. After that is accounted for, places like France > and > > Switzerland come out the dominant (highest frequency) place. But England, > > and even Norway, still have a significant frequency of I2.0110*. (In > > contrast, the I2.00* branch for example, has a very low frequency in > > England, and a high frequency in Eastern Europe.) > > > > Terry > > > > 2012/2/23 Daniel Suárez Díaz <dsuarez_job@hotmail.com> > > > > > > > > Terry, > > > > > > Thank you for share your work. Let me make you some questions about the > > > branch I2.0110*. These branch correspon to the I2ab2 haplogroup and > > > acording to the studies of Hans de Beule, the present day distribution > is > > > limited to the Upper Rhine and the British Isles. Also in the I2ab2 > FDNA > > > project, the most of the samples are British. However in your I2.0110* > map > > > with 63 samples, the major figures correspon to France and Switzerland. > > > > > > Therefore my questions are: > > > > > > - How many French and Swiss samples have you used? > > > > > > - Which are your sources? > > > > > > - And if the sources are public, Have you overweighted them or in other > > > words, Have you underweighted the British samples because many of them > are > > > the same family? > > > > > > Thanks in advance. > > > > > > Daniel Suarez. > > > > > > > > > > > > > > > > Date: Thu, 23 Feb 2012 01:27:04 +1100 > > > > From: tdrobb@gmail.com > > > > To: Y-DNA-HAPLOGROUP-I@rootsweb.com > > > > Subject: [yDNAhgI] STR Branches, SNP's, Geographic Origin > > > > > > > > Often people ask about what SNP's they should look at testing, given > > > their > > > > particular STR values. In addition, people sometimes ask what can be > said > > > > about their ancestral geographic origin based again on their STR > values. > > > > > > > > One can in fact answer those types of questions in an objective > sense. > > > > > > > > The methodology I used is quite simple. First, I collected as much > STR > > > > haplotype data as I could (I got over eight thousand I1 and I2 > samples > > > from > > > > various FTDNA Projects and from the Ysearch database). Next, I > organised > > > > that large STR dataset into "hierarchical clusters" using a standard > > > > mathematical technique. Finally, I counted up the number of SNP > alleles > > > > (either positive or negative) that occurred within each cluster, now > > > called > > > > an STR "Branch", and I displayed the results; and I also plotted > maps of > > > > the frequency distribution of the most-distant-known-ancestor of > each STR > > > > Branch member across all regions of Europe. > > > > > > > > The results are very informative. > > > > > > > > The answer to many questions just depend on knowing your STR "Branch > > > Code". > > > > In the system I use, the Branch Code is essentially just a way of > > > labelling > > > > the STR branches in the computed hierarchical tree, and it is very > > > similar > > > > to the "Henry System" used in genealogy for numbering the > descendants of > > > an > > > > ancestor. > > > > > > > > To determine your very own STR Branch Code in the system I use, see: > > > > http://www.goggo.com/cgi-bin/branchFind.cgi > > > > > > > > Enter you FTDNA Kit Number or Ysearch ID, and (hopefully) it will > return > > > > your STR Branch Code. (Only works for y-haplogroups I1 and I2, and > 67 or > > > > more STR markers at present.) > > > > > > > > Follow the links to see what your STR Branch Code can tell you. > > > > > > > > Terry > > > > > > > > ------------------------------- > > > > To unsubscribe from the list, please send an email to > > > Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' > > > without the quotes in the subject and the body of the message > > > > > > > > > ------------------------------- > > > To unsubscribe from the list, please send an email to > > > Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' > > > without the quotes in the subject and the body of the message > > > > > > > ------------------------------- > > To unsubscribe from the list, please send an email to > Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' > without the quotes in the subject and the body of the message > > > ------------------------------- > To unsubscribe from the list, please send an email to > Y-DNA-HAPLOGROUP-I-request@rootsweb.com with the word 'unsubscribe' > without the quotes in the subject and the body of the message >