I think my point was misunderstood. It is maintained that the AAV values are represented by set 1a values. My point is that I don't think this represents the ancestral values for group A since none of the participants outside the US match them. Most of the American participants have Colonial American roots with presumed, but yet undetermined connections, to each other. Their DNA should be close to each other and you should get a number of exact or close matches. This is the case. Just because we over sampled this large family does not mean their AAV is the true ancestral values. Let me propose a hypothetical scenario which may highlight my point better for you and generate a fresh look at the data. Suppose DNA testing was banned in the US and we did not have a DNA tests from participants in the US. Thus, our project only consisted of those Kincaids who hailed from outside the US. The AAV for them is then different. Unless we get paticipants from outside the US with values lining up with set 1a, then the true ancestral values is not determined but not likely that of set 1a. Best wishes! Peter P.S. I greatly respect the work of Sue and Don. I am just debating one of the assertions made given my perspective of the data. ----- Original Message ----- From: Sue Liedtke To: kincaid@rootsweb.com Sent: Tuesday, December 09, 2008 4:56 PM Subject: Re: [KINCAID] 67 marker DNA test analysis & info on TheGatheringofthe Clans in Scotland It is not the "exact" matches across all the markers that determine the AAV but rather the results so far at each marker, whether the same result appears across sets and whether variances are up and down from the most frequent value. Concentrating on only "exact" matches across the entire test to draw conclusions is extremely limiting when clearly there are scattered mutations that have occured within a known time period. If I have counted correctly (and I may not have) 15 of the first 37 markers do not show any mutation at all. The AAV value at those markers is unquestionable. 8 markers have a single incident of mutation. The AAV for these markers also cannot be questioned. To my way of thinking the single mutation would be of recent origin i.e. the common ancestor with others within the set did not have that mutation. It can be disregarded unless someone else in the same set shows up with it. 7 markers have 2-5 individuals with mutations. The mutations are scattered without an apparent pattern and are not always the same value (some are up and some are down). Here again I believe that they are downline from the common ancestor for the set they are in. That leaves 7 markers which have some concentration of participants having a given mutation value. Mutations in the CDY complex (markers 34 & 35) are common, scattered across sets or grouped within sets. There appears to have been multiple incidences of parallel mutations both up and down. The values chosen for the AAV appear more frequently, have values up and down away from them, and appear in every set (except set 4 in the case of CDYa). 3 markers (437, 460, 456) are subset defining. There is a definate grouping within a set that can be descerned using those markers. The overwhelming majority of individuals in both sets have the assigned AAV for these markers. . Marker 4 (391) appears to mutate readily from a 12 to an 11 with several proven parallel mutations. Although the mutation appears as a subset defining marker by default, its history makes basing probable closer relationship between those who have it iffy. The AAV is undoubtedly 12 as that value appears across all subsets and has not in any case been shown to be the result of parallel mutation. That leaves us with marker 15 (459b) with 2 possible Apparent Ancestral Values. Eventually we may have sufficient data and documentation to show whether the 10 or the 9 is the correct AAV but perhaps we will never know. The chosen AAV at this marker is a guess. 10 is the most frequent result across ALL other Rb1 groups in the project. This marker appears to be quite stable at 10. The 10 appears in a descendent of James and Helen Scott who married in Scotland in 1669. This family did not leave Scotland until well after the early colonial immigrants arrived on this side of the Atlantic. We can have as arguements as we would like about how the first Kincaid ancestor had a mutation to 9 and a branch of his family then mutated back to the RB1 value of 10. Given the stability of the value 9 among non-RB1 groups makes this less likely. There is a possibility that a parallel mutation to 9 did take place in the case of 49289. Documentation suggests an A-1 ancestor. Hopefully we will have another test for this line that will either confirm the value or suggest that a parallel mutation took place. Sue Liedtke ----- Original Message ----- From: "Peter A. Kincaid" <7kincaid@nb.sympatico.ca> To: <kincaid@rootsweb.com> Sent: Monday, December 08, 2008 12:27 PM Subject: Re: [KINCAID] 67 marker DNA test analysis & info on The Gatheringofthe Clans in Scotland > Correct me if I am wrong but none of the Group A > participants with roots outside United States match > set 1a marker values. Is this not clearly relevant? > > Peter

Peter, 36 of the first 37 markers have the selected marker AAV prevalent in both sets 1 and 2. I have conceded that the AAV for marker 15 is a guess. There are ancestors who did not immigrate into Colonial America (both US and Canada) in both sets. In A1: 94749, 2617, and 23547. In A2: 2519. There are undoubtedly others whose ancestors arrived after the Colonial periods (I am using 1867 as the end to the Canadian Colonial period. Is this correct?) but I don't know their lineages well enough to include them in the list. That there are descendents of US Colonial immigrants who happen to match the AAV exactly on these 36 markers is totally irrelevant to the AAV selection for each marker. The AAV selection is outlined in the trailer I purposely left on. Again note that the selections except on marker 15 are based on values occuring in both sets. If you removed all US participants Group A would shrink to the point where there is not enough data to predict an AAV on several markers but the AAV for the vast majority of markers would remain the same. Set 4 would dominate in number of participants (who all seem to trace to a single historic ancestor) but may not have been placed in a group with the others because of the number of variances. Set 2b would disappear entirely. Set 2a would have one participant. Set 1c may have 1 participant. Set 1b would disappear. Set 1a would have 2 participants. Groups B & C1 (both groups seem to have a possible common ancestor for the majority of participants within historic times) would have participant weight equal to the new Group A but their values would not affect the AAV for Group A. The whole project would shrink to the point where it would be quite difficult to draw any conclusions at all. Note: I did not look up current residences of the actual participants so the numbers above are guesses based on memory and chart info for furthest ancestor but will serve to illustrate my point. Exactly what is your point. Sue Liedtke >I think my point was misunderstood. It is maintained > that the AAV values are represented by set 1a values. > My point is that I don't think this represents the > ancestral values for group A since none of the participants > outside the US match them. Most of the American > participants have Colonial American roots with presumed, > but yet undetermined connections, to each other. Their > DNA should be close to each other and you should > get a number of exact or close matches. This is the > case. Just because we over sampled this large family > does not mean their AAV is the true ancestral values. > > Let me propose a hypothetical scenario which may > highlight my point better for you and generate a > fresh look at the data. Suppose DNA testing was > banned in the US and we did not have a DNA tests > from participants in the US. Thus, our project only > consisted of those Kincaids who hailed from outside > the US. The AAV for them is then different. > Unless we get paticipants from outside the US with > values lining up with set 1a, then the true ancestral > values is not determined but not likely that of set 1a. > > Best wishes! > > Peter > > P.S. I greatly respect the work of Sue and Don. I am > just debating one of the assertions made given my > perspective of the data. > > > > It is not the "exact" matches across all the markers that determine the > AAV > but rather the results so far at each marker, whether the same result > appears across sets and whether variances are up and down from the most > frequent value. Concentrating on only "exact" matches across the entire > test to draw conclusions is extremely limiting when clearly there are > scattered mutations that have occured within a known time period. > > If I have counted correctly (and I may not have) 15 of the first 37 > markers > do not show any mutation at all. The AAV value at those markers is > unquestionable. > > 8 markers have a single incident of mutation. The AAV for these markers > also > cannot be questioned. To my way of thinking the single mutation would be > of > recent origin i.e. the common ancestor with others within the set did not > have that mutation. It can be disregarded unless someone else in the same > set shows up with it. > > 7 markers have 2-5 individuals with mutations. The mutations are > scattered > without an apparent pattern and are not always the same value (some are > up > and some are down). Here again I believe that they are downline from the > common ancestor for the set they are in. > > That leaves 7 markers which have some concentration of participants > having a > given mutation value. > > Mutations in the CDY complex (markers 34 & 35) are common, scattered > across > sets or grouped within sets. There appears to have been multiple > incidences > of parallel mutations both up and down. The values chosen for the AAV > appear > more frequently, have values up and down away from them, and appear in > every > set (except set 4 in the case of CDYa). > > 3 markers (437, 460, 456) are subset defining. There is a definate > grouping > within a set that can be descerned using those markers. The overwhelming > majority of individuals in both sets have the assigned AAV for these > markers. . > > Marker 4 (391) appears to mutate readily from a 12 to an 11 with several > proven parallel mutations. Although the mutation appears as a subset > defining marker by default, its history makes basing probable closer > relationship between those who have it iffy. The AAV is undoubtedly 12 as > that value appears across all subsets and has not in any case been shown > to > be the result of parallel mutation. > > That leaves us with marker 15 (459b) with 2 possible Apparent Ancestral > Values. Eventually we may have sufficient data and documentation to show > whether the 10 or the 9 is the correct AAV but perhaps we will never > know. > The chosen AAV at this marker is a guess. 10 is the most frequent result > across ALL other Rb1 groups in the project. This marker appears to be > quite > stable at 10. The 10 appears in a descendent of James and Helen Scott who > married in Scotland in 1669. This family did not leave Scotland until > well > after the early colonial immigrants arrived on this side of the Atlantic. > We > can have as arguements as we would like about how the first Kincaid > ancestor > had a mutation to 9 and a branch of his family then mutated back to the > RB1 > value of 10. Given the stability of the value 9 among non-RB1 groups > makes > this less likely. > > There is a possibility that a parallel mutation to 9 did take place in > the > case of 49289. Documentation suggests an A-1 ancestor. Hopefully we will > have another test for this line that will either confirm the value or > suggest that a parallel mutation took place. > > Sue Liedtke >