Naming of novel alleles with large numbers of SNPs?

[scharch]

Is this a bug or did you decide that adding 7 SNPs to the allele name was too unwieldy?

 hv2-2601              1089 counts
             position   ratio       (one piece / two pieces)              0          1          2          3          4          5          6          7          8          9         10         11         12         13         14         15         16         17      
    7 snps
                  93     7.11             5.29 / 0.74                   0 / 268    0 / 234    1 / 104    0 / 81     1 / 49     1 / 35     2 / 30    15 / 43    10 / 34    12 / 30     5 / 23    12 / 28    14 / 28     5 / 14    10 / 21     4 / 11     3 / 11              
                 166     4.77             3.09 / 0.65                   0 / 268    0 / 234    0 / 104    0 / 81     0 / 49     3 / 35     0 / 30    11 / 43     7 / 34     9 / 30     3 / 23     9 / 28     5 / 28     4 / 14     7 / 21     3 / 11     3 / 11              
                 110     4.44             3.28 / 0.74                   0 / 268    2 / 234    0 / 104    0 / 81     1 / 49     0 / 35     0 / 30    13 / 43     8 / 34     9 / 30     3 / 23    10 / 28     5 / 28     5 / 14     7 / 21     2 / 11     3 / 11              
                 248     4.33             2.78 / 0.64                   0 / 268    0 / 234    0 / 104    4 / 80     1 / 49     1 / 35     0 / 30    12 / 43     8 / 34     8 / 30     4 / 23    10 / 28     5 / 28     4 / 14     7 / 21     2 / 11     3 / 11              
                 247     4.19             4.09 / 0.98                   0 / 267    0 / 234    1 / 104    0 / 81     0 / 48     0 / 35     1 / 30    13 / 43     8 / 34    10 / 30     3 / 23    11 / 28     6 / 28     7 / 14     7 / 21     2 / 11     4 / 11              
                  84     3.56             2.51 / 0.71                   0 / 268    0 / 234    1 / 104    2 / 81     0 / 49     1 / 35     1 / 30    12 / 43     7 / 34     9 / 30     3 / 23     9 / 28     4 / 28     4 / 14     7 / 21     2 / 11     3 / 11              
                  88     3.51             2.68 / 0.76                   0 / 268    2 / 234    1 / 104    0 / 81     1 / 49     1 / 35     1 / 30    13 / 43     7 / 34     8 / 30     4 / 23    10 / 28     4 / 28     4 / 14     8 / 21     2 / 11     4 / 11              
   snps       min ratio
    7          3.51     candidate 
        no equivalent gene for hv2-2601+32329       nearest is hv2-2601 (7 snps, 0 indels):
            CAGGTCACCTTGAAGGAGTCTGGTCCTGTGCTGGTGAAACCCACAGAGACCCTCACGCTGACCTGCACCGTCTCTGGGTTCTCACTCAGCAATGCTAGAATGGGTGTGAGCTGGATCCGTCAGCCCCCAGGGAAGGCCCTGGAGTGGCTTGCACACATTTTTTCGAATGACGAAAAATCCTACAGCACATCTCTGAAGAGCAGGCTCACCATCTCCAAGGACACCTCCAAAAGCCAGGTGGTCCTTACCATGACCA
            CAGGTCACCTTGAAGGAGTCTGGTCCTGTGCTGGTGAAACCCACAGAGACCCTCACGCTGACCTGCACCGTCTCTGGGTTCTCATTCACCAATCCTAGAATGGGTGTGAGTTGGATCCGTCAGCCCCCAGGGAAGGCCCTGGAGTGGCTTGCACACATTTTTTCGAGTGACGAAAAATCCTACAGCACATCTCTGAAGAGCAGGCTCACCATCTCCAAGGACACCTCCAAAAGCCAGGTGGTCCTTATGATGACCA
  new hv2-2601+32329 separated from hv2-2601 by 7 snps at:   84 (C --> T)  88 (G --> C)  93 (G --> C)  110 (C --> T)  166 (A --> G)  247 (C --> T)  248 (C --> G)

[psathyrella]

Not a bug! But open to arguments that it's a questionable design decision ;-). The cutoff is five SNPs (or any number of indels), after that it instead uses a hash() of the sequence. Naming ends up being quite seriously complicated, and has quite a bit of inescapable arbitrariness (around which gene should be the template, whether the index is zero- (partis) or one-based (others)), so I would recommend viewing any gene name as a frequently-useful shorthand that contains some amount of information about the gene's relationship to other genes, but the only really reliable identifier is the sequence.

[scharch]

Thanks