for g in gsets[ii]:
counts[g] = counts.get(g, 0) + 1
mx = max(counts.values())
genesets[ii] = set(
[x for x, y in counts.iteritems() if y == mx])
print 'empty common genes:', ii, 'clone_size:', clone_size, 'mx-genecount', mx, 'newgeneset:', genesets[
ii], em
for genes, segtype in zip(genesets, segtypes_lowercase):
assert genes
tag = segtype + '_genes'
assert tag in outl # should already be there, now over-writing
outl[tag] = ';'.join(sorted(genes))
## update reps
reps = sorted(set((util.get_rep(x, organism) for x in genes)))
tag = segtype + '_reps'
assert tag in outl # should already be there, now over-writing
outl[tag] = ';'.join(reps)
## update countreps
countreps = sorted(
set((util.get_mm1_rep_gene_for_counting(x, organism)
for x in genes)))
tag = segtype + '_countreps'
assert tag in outl # should already be there, now over-writing
outl[tag] = ';'.join(countreps)
total_clones += 1
out.write(make_tsv_line(outl, outfields, '-') + '\n')
def reconstruct_field_from_data(field, l, organism):
try:
if field in l:
return l[field]
if field == 'subject':
if 'mouse' in l:
return l['mouse']
elif field.endswith('_prob'):
if prob_warning not in warnings:
print prob_warning
warnings.add(prob_warning)
return 1.0
elif field == 'clone_size':
if clone_size_warning not in warnings:
print clone_size_warning
warnings.add(clone_size_warning)
return 1
elif field[:3] in ['va_', 'ja_', 'vb_',
'jb_']: ## these are all pretty similar
prefix = field[:3]
tag = field[3:]
#print 'prefix:',prefix,tag,l.keys()
if tag == 'gene':
# only one place to get this
genes_field = prefix + 'genes'
if genes_field not in l:
return None
genes = l[genes_field].split(listsep)
#print 'genes:',genes
return sorted(genes)[0]
elif tag == 'genes':
if prefix + 'blast_hits' in l:
hits = l[prefix + 'blast_hits']
if hits == '-': #failed to find any hits
return None
else:
return listsep.join(sorted(util.get_top_genes(hits)))
elif prefix + 'gene' in l: ## just take the one gene we know
return l[prefix + 'gene']
elif tag == 'rep':
if prefix + 'gene' in l:
if "*" in l[prefix + 'gene']:
return util.get_rep(l[prefix + 'gene'], organism)
else:
if rep_warning not in warnings:
print rep_warning
warnings.add(rep_warning)
return l[prefix + 'gene']
elif tag == 'reps':
## we should already have hit 'genes' in the list of fields we are trying to fill !!!
if "*" in l[prefix + 'gene']:
return listsep.join(
sorted((util.get_rep(x, organism)
for x in l[prefix + 'genes'].split(listsep))))
else:
if rep_warning not in warnings:
print rep_warning
warnings.add(rep_warning)
return l[prefix + 'gene']
elif tag == 'countreps':
if "*" in l[prefix + 'gene']:
return listsep.join(
sorted(
util.countreps_from_genes(
l[prefix + 'genes'].split(listsep), organism)))
else:
if rep_warning not in warnings:
print rep_warning
warnings.add(rep_warning)
return l[prefix + 'gene']
elif field.endswith('_quals'):
seqfield = field[:5] + '_nucseq'
if seqfield not in l:
return None
return '.'.join(['60'] * len(l[seqfield]))
except Exception as inst: ## this is not the best way to handle it...
print 'Hit an exception trying to get field {} from line'.format(
field), inst
print 'Failed to reconstruct {} from the input fields: {}'\
.format( field, ' '.join( sorted( l.keys() ) ) )
return None
def __init__(self, organism, va_genes, vb_genes, cdr3a, cdr3b):
self.va_reps = frozenset((util.get_rep(x, organism) for x in va_genes))
self.vb_reps = frozenset((util.get_rep(x, organism) for x in vb_genes))
self.cdr3a = cdr3a[:]
self.cdr3b = cdr3b[:]