Recurrent Mutations

[Mike McG]

With the significant number of Y positions now being tested by FGC, BigY or other “NGS” type of technology, it would seem the possible occurrence (or discovery) of recurrent mutations might increase dramatically every year as more individuals are tested. When single SNPs were tested with Sanger sequencing it was unlikely a test would be performed for a SNP in a different Haplogroup; but initially with chip based sequencing such as Geno, and now with “NGS” type sequencing, there are so many more SNPs tested in multiple Haplogroups that the discovery of recurrent mutations (forward, back or deletion) is bound to occur.

Even though these recurrent SNPs are unreliable as a single terminal SNP test they may still be significant in determining branches or in finding closer Y relatives. I hope that (with in reason) ISOGG continues to use the “.n” or “_n” SNP extensions, that the SNP DNA companies continue to offer Sanger sequencing of them, and that these SNPs continue to be incorporated in the Haplotrees where appropriate.

Mike McG

[Cofgene]

Even though these recurrent SNPs are unreliable as a single terminal SNP test they may still be significant in determining branches or in finding closer Y relatives. I hope that (with in reason) ISOGG continues to use the “.n” or “_n” SNP extensions, that the SNP DNA companies continue to offer Sanger sequencing of them, and that these SNPs continue to be incorporated in the Haplotrees where appropriate.

Mike McG

Recurrent SNPs should be considered unreliable as a terminal SNP if they flip within the same haplogroup branch and not whether they have multiple occurrences across the y-haplogroup tree.. The measure for a terminal SNP should be stability within the context that it is found. Testing will be available for most recurrent SNPs since the odds are good that their position will be covered by sequencing efforts for other nearby SNPs.

[Huntergatherer1066]

I think the particularly reason why they recurrent is important. If they are just random occurences of convergent evolution, that's fine, but if the supposed SNP is in or adjacent an STR or located in a psuedoautosomal region then you can have problems, especially when building phylogenetic trees.