MakeDb assigning failed germlines to None
Issue #140
resolved
MakeDb is assigning None
to GERMLINE_IMGT
when a failure during germline reconstruction occurs. This should be an empty string instead.
Comments (9)
-
reporter -
-
assigned issue to
-
assigned issue to
-
@javh I haven't been able to reproduce this in the example (S43) database. Any recommendations for data to reproduce this error?
-
reporter Change the values of one of the positional fields in the igblast file (reference start/end, etc). That should force it to fail.
-
Got it to work for igblast based on your suggestion. Think that's sufficient testing or do we need to try it on other input formats?
-
reporter That should be sufficient.
-
- changed status to resolved
changed line 589 of Receptor.py to:
# Nucleotide sequence type conversion @staticmethod def nucleotide(v, deparse=False): if not deparse: try: #return '' if v in ('NA', 'None') else Seq(v, IUPAC.ambiguous_dna).upper() return '' if v in ('NA', 'None') else v.upper() except: return '' else: return '' if v in ('NA', 'None', None) else str(v)
-
Also changed
# Sequence type conversion @staticmethod def aminoacid(v, deparse=False): if not deparse: try: #return '' if v in ('NA', 'None') else Seq(v, IUPAC.extended_protein).upper() return '' if v in ('NA', 'None') else v.upper() except: return '' else: return '' if v in ('NA', 'None', None) else str(v)
-
Both changes are in commit 3e7b3ef
- Log in to comment
May be specific to MakeDb-igblast.
First thing to check is that Receptor.ReceptorData.nucleotide (and
Receptor.ReceptorData.aminoacid
) properly convertsNone
to an empty string whendeparse=True
. Changing the tuple from('NA', 'None')
to('NA', 'None', None)
may be sufficient (need to test).