getSegment doesn't work for C-regions

[Jason Vander Heiden]

The allele, gene and family regexs don’t behave as expected for C region genes (IGHD specifically):

> countGenes(ExampleDb, gene="c_call", mode="allele")
# A tibble: 4 x 3
  gene  seq_count seq_freq
  <chr>     <int>    <dbl>
1 IGHM        719    0.360
2 IGHG        650    0.325
3 IGHA        372    0.186
4 IGH         259    0.130

> countGenes(ExampleDb, gene="c_call", mode="asis")
# A tibble: 4 x 3
  gene  seq_count seq_freq
  <chr>     <int>    <dbl>
1 IGHM        719    0.360
2 IGHG        650    0.325
3 IGHA        372    0.186
4 IGHD        259    0.130

This is probably fixable by changing (IG[HLK]|TR[ABGD])[VDJ] to (IG[HKL][VDJADGEMC]|TR[ABGD])[VDJC]. This should also work at the gene/allele level; eg, “IGHA1*01”, “IGHG2”, “IGKC”, "IGLC1", "TRBC2", etc.

See also #86.

CC: @Scott Christley

[Edel Aron]

So it seems like in addition to needing to updating the regex, the strip_d function in getSegment (which countGenes indirectly calls) being set to TRUE meant that the IGHD's weren't getting counted properly since they were automatically getting converted to IGH's.

[Edel Aron]

We decided to update strip_d instead of altering countGenes' strip_d call so that it won't strip D's when they refer to isotypes instead of duplicates (see this commit).

[Edel Aron]

The updated regex and strip_d function seems to have done the trick