James,   (with a copy to the Pace list as this applies to one of the defining markers between Groups 3a and 3b)   Bob Stafford's reply to you (James) on GenForum seems to have been misunderstood.  There is no recLOH at marker 385 in the Pace DNA project.   Bob Stafford, a long-term project administrator, wrote: http://genforum.genealogy.com/dna/messages/3540.html "For multi-copy markers (those labeled a, b, c, etc.), another mechanism, called a conversion (or recLoh), is possible. It is suspected when there is a multi-step mutation and a doubling of one marker. For example, in your case, if one branch were (12,14) on 385 and the other (12, 12) or (14,14), a conversion is another possible explanation." ----   There is no recLOH (recombinational loss of heterozygosity)that occurred in the Pace DNA project on marker DYS 385.    Group 3a has the value of 11-14 and Group 3b has the value of 11-12.    Those values are not indicative of recLOH - there is no doubling of one of the marker values.   When recLOH occurs, one of the groups would have had the results go to a double value.   One of the sets of results would be 11-11 or 14-14 or 12-12.  That is not the case in the Pace DNA project - there is no doubling of the values - just a two-step mutation.   Even if there were recLOH, and there isn't, it wouldn't change the close relationship of the two groups.   Another important reply from Bob Stafford to James about the Pace DNA project 2-step mutation at marker 385b follows: http://genforum.genealogy.com/dna/messages/3538.html "Yes, multi-step mutations are observed in father-son pairs. In the usual genealogical time frame, it is more likely to be a single event (with the possible exception of very fast markers such as CDY). It is best to count the number of matching markers when comparing two people of the same surname. An advantage is that two-step mutations are unlikely enough that it is a good branch-defining mutation, very unlikely to occur in another branch (parallel mutation)." ---       This reply of Bob Stafford's is exactly what we are seeing in the Pace DNA project.  We have a 2-step mutation that occurred between Groups 3a and 3b at marker 385b.  As FTDNA indicated to Roy early in the Pace DNA project, this 2-step mutation is rare and even though there is a 2-step mutation, Groups 3a and 3b are closely related.   A 2-step mutation does occur between father and son.   We now also have 67 marker results for both Groups 3a and 3b, and there really are only the 2 significant DNA changes between the two groups out of 67 markers tested - the rare 2-step mutation at marker 385b and the marker change at 449, which is a marker that has a good mutation rate making it a marker that helps separate branches of a same family in other surname projects as well.  The fact that between the two groups there is basically a 65/67 marker match (and even a higher match if you include the tested markers from other labs), also very strongly suggests these two groups are closely related.  There are some other DNA changes among different members of the 2 groups, but they are not significant on a group basis - only 385b and 449 are markers used for separating Groups 3a and 3b.   As Bob Stafford indicated, these rare two-step (or multi-step) mutations are "good branch defining mutation[s], very unlikely to occur in another branch (parallel mutation)."   That is, the 2-step mutation at marker 385b helps us separate different branches of the same family. Rebecca Christensen --- On Tue, 1/5/10, Roy Johnson <royj@webster.edu> wrote: From: Roy Johnson <royj@webster.edu> Subject: RE: reLOH mutations To: "'james pace'" <mogone.dan@hotmail.com> Cc: "'Rebecca Christensen'" <rchristen@sbcglobal.net> Date: Tuesday, January 5, 2010, 8:50 PM I saved your email for later response. I need to do more reading on slip strand and reLOH mucations.   Rebecca Christensen is our best expert on DNA and here is her conclusion: -------- As an aside, in case someone is wondering why I indicated Group 3b changed from Group 3a,  the "defining" markers for Group 3a also match the London participant who we know the common ancestor with Group 3a is further in the past - so the earlier shared Pace ancestor also had the Group 3a "defining" marker results. Group 3a also matches the results at the Group 3 "defining" markers for the larger population group Haplogroup R1b that Groups 3a and 3b belong to.  So it is Group 3b that has had the change from the ancient population group as a whole, besides being different from the results of the shared common ancestor between Group 3a and the London participant.  So, Group 3a results are the "ancestral" Pace DNA results for Group 3, while Group 3b has had the change from the "ancestral" Pace DNA results.   So the reason for the conclusion is the fact that 3a seems to reach further back into England because of the match with the London Pace who is for sure not a descendent of American Paces.   Maybe Rebecca would want to discuss or explain this further.   Roy Johnson     From: james pace [mailto:mogone.dan@hotmail.com] Sent: Tuesday, December 22, 2009 8:31 AM To: dna@pacesociety.com Subject: reLOH mutations   It seems to be assumed that group 3b in your study would be a mutation from the 3a group. This assumption maybe incorrect. Double mutations, as occurring at site 385b, are thought to be more frequently a result of recLOH than slip strand. reLOH mutation would explain the change in allele from 12 to 14.   Hotmail: Trusted email with powerful SPAM protection. Sign up now.