In a message dated 10/27/2011 5:38:05 A.M. Central Daylight Time, pabloburns@comcast.net writes: Did someone post that M222 must be relatively young for a SNP because the STR values do not show wide variance? I may be misremembering, but I made a comparison chart for the 20 predicted (by me based on STR indicators) and tested M222s in my project who have tested to 67 markers. I found a great many GDs of 18, 19, and 20 among them. As a follow-up exercise, I made a smaller chart to compare five NWI members who live in different areas of Ireland and whose known ancestors were from the same area. I added one USA resident whose EKA cannot be traced to Scotland, but he was called "that canny Scot" by no less than George Washington whom he bested in a land sale. "Kildare" and "Scotland" are confirmed M222. Their GD comparisons came out as follows (numbers across top in same order as descending): 117643 Donegal 0 7 10 10 14 13 36715 Sligo 7 0 12 13 17 19 149181 Roscommon 10 12 0 12 13 17 92019 Monaghan 10 13 9 0 13 14 145320 Kildare 14 17 11 14 0 20 20413 Scotland? 13 19 17 14 20 0 All I can see in this is that the Scotland entry shows the greatest variance, and therefore may be the oldest. That's an interesting question Paul and one I've wondered about for years. I've posted this before but it's relevant to this discussion. "Moffat: Because Nial of the Nine hostages is a figure of some historical substance around 400,500 AD. that means people with his marker in Scotland have moved from Ireland and that's there's been a substantial migration. Wilson: I think that is the case. Wilson: I must say it's not entirely proven yet because we would need to see higher diversity in these lineages, more different subtypes in Ireland than we do in Scotland and that's not clear yet but I'm convinced with more data we willl see that and the case will be closed." This is a short extract of the conversation between Moffat and Wilson. They're talking about diversity in STRs as a sign of an older population. Here the assumption is that M222 originated in Ireland and from there spread to Scotland. But as Wilson (of EthnoAncestry) says it's not proven yet. My problem with this is how do you measure diversity in STRs? Is it, as Wilson states, a matter of more different subtypes in one location? I'm not sure how that would be quantified. Is genetic distance alone a measure of diversity? What qualifies as a subtype? Something as simple as 13-25-14-11 vs. 13-25-14-10? One of the more interesting yet so far unexplained aspects of Bill Howard's analysis of M222 is his finding of a bi-modality within M222. I've seen the graphs. It's there. I trust his statistical wizardry even if I don't understand it myself. One idea he came up with is the bi-modality could reflect a division between descendants of Niall and of those who are pre-Niall including the Connachta. But it seems to me that if such a bi-modality is obvious statistically then it should also be evident in the STRs themselves. Yet I never really got an answer from him on that - or about which samples fell into which side of the bi-modality. The charts he made look like a two humped camel - fairly evenly divided as I recall. I recently heard from a Dutch M222 and a German M222. Somewhat earlier from a Swedish M222. Those samples are still out there and unexplained. They can't all be NPEs. John

[My problem with this is how do you measure diversity in STRs? Is it, as Wilson states, a matter of more different subtypes in one location? I'm not sure how that would be quantified. Is genetic distance alone a measure of diversity? What qualifies as a subtype? Something as simple as 13-25-14-11 vs. 13-25-14-10?] The classical statistical way would be mean marker variance. That's what KN uses. The main advantage of variance is that it can be shown mathematically that the expected value of the variance after G generations is EV = mG where EV=expected variance, m=mutation rate, G=number of generations. There are other measures, equally valid, though. Examples are mean pairwise GD, observed proportion on modal (the higher the proportion on modal the lower the diversity), or even mean number of matches over all pairwise comparisons. These other measures are best used in comparison with other haplogroups or sub-clades, because unlike with variance, they don't give a direct answer for an estimate of G. So we can compare M222+ with about 90% on modal over both 67 markers and 111 markers With L21+ which is a touch below 80% on modal. In the context that Wilson uses 'subtypes', I think he means 'sub-clades', or 'branches'. Sandy -----Original Message----- From: dna-r1b1c7-bounces@rootsweb.com [mailto:dna-r1b1c7-bounces@rootsweb.com] On Behalf Of Lochlan@aol.com Sent: 28 October 2011 00:52 To: dna-r1b1c7@rootsweb.com Subject: Re: [R-M222] Age of M222 In a message dated 10/27/2011 5:38:05 A.M. Central Daylight Time, pabloburns@comcast.net writes: Did someone post that M222 must be relatively young for a SNP because the STR values do not show wide variance? I may be misremembering, but I made a comparison chart for the 20 predicted (by me based on STR indicators) and tested M222s in my project who have tested to 67 markers. I found a great many GDs of 18, 19, and 20 among them. As a follow-up exercise, I made a smaller chart to compare five NWI members who live in different areas of Ireland and whose known ancestors were from the same area. I added one USA resident whose EKA cannot be traced to Scotland, but he was called "that canny Scot" by no less than George Washington whom he bested in a land sale. "Kildare" and "Scotland" are confirmed M222. Their GD comparisons came out as follows (numbers across top in same order as descending): 117643 Donegal 0 7 10 10 14 13 36715 Sligo 7 0 12 13 17 19 149181 Roscommon 10 12 0 12 13 17 92019 Monaghan 10 13 9 0 13 14 145320 Kildare 14 17 11 14 0 20 20413 Scotland? 13 19 17 14 20 0 All I can see in this is that the Scotland entry shows the greatest variance, and therefore may be the oldest. That's an interesting question Paul and one I've wondered about for years. I've posted this before but it's relevant to this discussion. "Moffat: Because Nial of the Nine hostages is a figure of some historical substance around 400,500 AD. that means people with his marker in Scotland have moved from Ireland and that's there's been a substantial migration. Wilson: I think that is the case. Wilson: I must say it's not entirely proven yet because we would need to see higher diversity in these lineages, more different subtypes in Ireland than we do in Scotland and that's not clear yet but I'm convinced with more data we willl see that and the case will be closed." This is a short extract of the conversation between Moffat and Wilson. They're talking about diversity in STRs as a sign of an older population. Here the assumption is that M222 originated in Ireland and from there spread to Scotland. But as Wilson (of EthnoAncestry) says it's not proven yet. My problem with this is how do you measure diversity in STRs? Is it, as Wilson states, a matter of more different subtypes in one location? I'm not sure how that would be quantified. Is genetic distance alone a measure of diversity? What qualifies as a subtype? Something as simple as 13-25-14-11 vs. 13-25-14-10? One of the more interesting yet so far unexplained aspects of Bill Howard's analysis of M222 is his finding of a bi-modality within M222. I've seen the graphs. It's there. I trust his statistical wizardry even if I don't understand it myself. One idea he came up with is the bi-modality could reflect a division between descendants of Niall and of those who are pre-Niall including the Connachta. But it seems to me that if such a bi-modality is obvious statistically then it should also be evident in the STRs themselves. Yet I never really got an answer from him on that - or about which samples fell into which side of the bi-modality. The charts he made look like a two humped camel - fairly evenly divided as I recall. I recently heard from a Dutch M222 and a German M222. Somewhat earlier from a Swedish M222. Those samples are still out there and unexplained. They can't all be NPEs. John R1b1c7 Research and Links: http://clanmaclochlainn.com/R1b1c7/ ------------------------------- To unsubscribe from the list, please send an email to DNA-R1B1C7-request@rootsweb.com with the word 'unsubscribe' without the quotes in the subject and the body of the message

I hope I have succeeded in not including everything in the thread in this post. I'm copying a bit below: *In a message dated 10/27/2011 5:38:05 A.M. Central Daylight Time,* *pabloburns@comcast.net writes:* * * *Did someone post that M222 must be relatively young for a SNP because the* *STR values do not show wide variance? I may be misremembering, but I made a* * comparison chart for the 20 predicted (by me based on STR indicators) and* *tested M222s in my project who have tested to 67 markers. I found a great* *many GDs of 18, 19, and 20 among them.* *...* and *On Thu, Oct 27, 2011 at 6:52 PM, <Lochlan@aol.com> wrote:* *This is a short extract of the conversation between Moffat and Wilson.* *They're talking about diversity in STRs as a sign of an older population. * *Here the assumption is that M222 originated in Ireland and from there spread* *to Scotland. But as Wilson (of EthnoAncestry) says it's not proven yet. * * * *My problem with this is how do you measure diversity in STRs? Is it, as* *Wilson states, a matter of more different subtypes in one location? I'm not* *sure how that would be quantified. Is genetic distance alone a measure* *of diversity? What qualifies as a subtype? Something as simple as* *13-25-14-11 vs. 13-25-14-10?* *...* I've been thinking about this lately because of the recent discussions of family mutations as well as the fact that in one of the charts available in the files of the R-L21+ Yahoo Group, my family representative (kit 161531) is listed as McCombs 461=12, 481<=23. The other McCombs in the R-M222 Haplogroup (kit 175471) might similarly be listed as McCombs 391=10, 461=12, 481<=23, as would his cousin (kit 182872). They'd be put in a separate group from kit 161531 even though the GD for 161531 and 175471 @ 67 is only 2. There used to be a chart with a link posted in the R-M222 Haplogroup Project pages that was a star of sorts - lines radiating outward. On this chart there were three markers that were given special significance and haplotypes with a particular value for those markers were highlighted in different colors. I can no longer find that chart and I can only recall two of the markers, DYS391 = 10 and DYS460 = 12. I don’t believe the 3rd mutation occurred in any of our R-M222 McComb(s). However, there is one McComb (kit 196033) whose DYS460 is the modal 11, rather than the 12 the other 3 McComb(s) have, so he'd probably be put into a 3rd group. Comparing 161531 to 196033 the GD is 2, and comparing 175471 to 196033 the GD is 4. The only way these kits would end up in the same group, is if they were first put into a group with other haplotypes that have the DYS481=22. It's an unusual back mutation (?) and in the 4th panel, so it's not something that anyone not connected to this particular group is likely to notice. Since most comparisons are made with 37 markers rather than 67, DYS481 isn't usually considered, but it defines a group that includes the East Lothian Duncan family, a segment of Group 01 McKee family, most of the M222 Group 8 Walkers, the 4 McComb(s) above, and a number of others. I've collected 27 haplotypes from individuals on the match list of 161531 (and the public Y-charts of his matches' projects) who have 22 at 481. Only 12 of them are in the R-M222 Haplogroup Project, but they're easy to spot by the purple highlighting of DYS481 in the colorized chart. (Unfortunately, no one with the surname Walker is among the 12 in the haplogroup project.) Few of these individuals have 111 marker results, but those who do still show up on each other's match lists at that level and have at least one more off-modal marker in common, DYS461=13. Only 2 of the 27 in my spreadsheet have 391=10 and they both descend from David McCombs b. 1793 in Abbeville SC, however 21 of the 27 have 460=12. Of the other 6 individuals, one (Walker) has 460=13, and 5 have 460=11 (McKie, Walker, McComb, Ferguson Trumble, and Burditt McKee). Neither of these markers appear to give me a way to subdivide the group; there is only one marker that does that - DYS444, also in the 4th panel and adjacent to 481. 15 of the 27 have the M222 modal 12 for DYS444, and 12 of the 27 have 13 there. The 12 haplotypes with 13 all belong to individuals named Walker, Woodruff or McComb(s), and no one with those names has 12. (There are a couple of other mutations that occur only in Woodruff and some individuals in the Walker family that could further subdivide the group.) It seems to me that there must be more groups like this one (and the others that were discussed recently) that have different signature markers and patterns of mutation within them that puts them outside the norm for R-M222. But perhaps the combined total of haplotypes from these groups is still so small that they would not affect the computations. Sharon >