def cdr3_translation(info):
naive_cdr3_seq = naive_cdr3(info)
naive_cdr3_seq = naive_cdr3_seq[3:len(naive_cdr3_seq) - 3]
if len(naive_cdr3_seq) % 3 != 0:
# print ' out of frame: adding %s' % ((3 - len(naive_cdr3_seq) % 3) * 'N')
naive_cdr3_seq += (3 - len(naive_cdr3_seq) % 3) * 'N'
return utils.ltranslate(naive_cdr3_seq)
def get_cdr3_title(self, annotation):
naive_cdr3_seq, _ = utils.subset_iseq(annotation, 0, restrict_to_region='cdr3')
title = ''
if len(naive_cdr3_seq) % 3 != 0:
# print ' out of frame: adding %s' % ((3 - len(naive_cdr3_seq) % 3) * 'N')
naive_cdr3_seq += (3 - len(naive_cdr3_seq) % 3) * 'N'
title += ' (out of frame)'
title = utils.ltranslate(naive_cdr3_seq) + title
return title
def increment_per_sequence_params(self, info, iseq):
""" increment parameters that differ for each sequence within the clonal family """
self.mute_total += 1
self.mfreqer.increment(info, iseq)
for nuke in utils.nukes:
self.counts['seq_content'][nuke] += info['seqs'][iseq].count(nuke)
# aa seq content stuff
nseq = info['seqs'][iseq]
if info['v_5p_del'] > 0:
nseq = info['v_5p_del'] * utils.ambig_base + nseq
if len(info['fv_insertion']) > 0:
nseq = nseq[len(info['fv_insertion']):]
if len(nseq) % 3 != 0:
nseq += utils.ambig_base * (
3 - (len(nseq) % 3)
) # I think I could replace this with the new utils.pad_nuc_seq()
aaseq = utils.ltranslate(nseq)
for aa in self.all_aa:
self.counts['seq_aa_content'][aa] += aaseq.count(aa)
def init_aa_stuff(self):
codons = itertools.product(
utils.nukes + ['N'], repeat=3
) # I cannot for the life of me find anything in Bio that will give me the list of amino acids, wtf, but I'm tired of googling, this will be fine
self.all_aa = set([utils.ltranslate(''.join(c)) for c in codons])