def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir)
perfplotter = PerformancePlotter(self.germline_seqs, self.args.plotdir, 'imgt')
# get sequence info that was passed to imgt
self.seqinfo = {}
with opener('r')(self.args.simfname) as simfile:
reader = csv.DictReader(simfile)
iline = 0
for line in reader:
if self.args.queries != None and line['unique_id'] not in self.args.queries:
continue
if len(re.findall('_[FP]', line['j_gene'])) > 0:
line['j_gene'] = line['j_gene'].replace(re.findall('_[FP]', line['j_gene'])[0], '')
self.seqinfo[line['unique_id']] = line
iline += 1
if self.args.n_queries > 0 and iline >= self.args.n_queries:
break
paragraphs, csv_info = None, None
if self.args.infname != None and '.html' in self.args.infname:
print 'reading', self.args.infname
with opener('r')(self.args.infname) as infile:
soup = BeautifulSoup(infile)
paragraphs = soup.find_all('pre')
summarydir = self.args.indir[ : self.args.indir.rfind('/')] # one directoy up from <indir>, which has the detailed per-sequence files
summary_fname = glob.glob(summarydir + '/1_Summary_*.txt')
assert len(summary_fname) == 1
summary_fname = summary_fname[0]
get_genes_to_skip(summary_fname, self.germline_seqs)
n_failed, n_skipped, n_total, n_not_found, n_found = 0, 0, 0, 0, 0
for unique_id in self.seqinfo:
if self.args.debug:
print unique_id,
imgtinfo = []
# print 'true'
# utils.print_reco_event(self.germline_seqs, self.seqinfo[unique_id])
if self.args.infname != None and '.html' in self.args.infname:
for pre in paragraphs: # NOTE this loops over everything an awful lot of times. Shouldn't really matter for now, though
if unique_id in pre.text:
imgtinfo.append(pre.text)
else:
n_total += 1
assert self.args.infname == None
infnames = glob.glob(self.args.indir + '/' + unique_id + '*')
assert len(infnames) <= 1
if len(infnames) != 1:
if self.args.debug:
print ' couldn\'t find it'
n_not_found += 1
continue
n_found += 1
with opener('r')(infnames[0]) as infile:
full_text = infile.read()
if len(re.findall('[123]. Alignment for [VDJ]-GENE', full_text)) < 3:
failregions = re.findall('No [VDJ]-GENE has been identified', full_text)
if self.args.debug and len(failregions) > 0:
print ' ', failregions
n_failed += 1
continue
# loop over the paragraphs I want
position = full_text.find(unique_id) # don't need this one
for ir in range(4):
position = full_text.find(unique_id, position+1)
pgraph = full_text[position : full_text.find('\n\n', position+1)]
if 'insertion(s) and/or deletion(s) which are not dealt in this release' in pgraph:
ir -= 1
continue
imgtinfo.append(pgraph) # query seq paragraph
if len(imgtinfo) == 0:
print '%s no info' % unique_id
continue
else:
if self.args.debug:
print ''
line = self.parse_query_text(unique_id, imgtinfo)
if 'skip_gene' in line:
# assert self.args.skip_missing_genes
n_skipped += 1
continue
try:
assert 'failed' not in line
joinparser.add_insertions(line, debug=self.args.debug)
joinparser.resolve_overlapping_matches(line, debug=False, germlines=self.germline_seqs)
except (AssertionError, KeyError):
print ' giving up'
n_failed += 1
perfplotter.add_partial_fail(self.seqinfo[unique_id], line)
# print ' perfplotter: not sure what to do with a fail'
continue
perfplotter.evaluate(self.seqinfo[unique_id], line)
if self.args.debug:
utils.print_reco_event(self.germline_seqs, self.seqinfo[unique_id], label='true:')
utils.print_reco_event(self.germline_seqs, line, label='inferred:')
perfplotter.plot()
print 'failed: %d / %d = %f' % (n_failed, n_total, float(n_failed) / n_total)
print 'skipped: %d / %d = %f' % (n_skipped, n_total, float(n_skipped) / n_total)
print ' ',
for g, n in genes_actually_skipped.items():
print ' %d %s' % (n, utils.color_gene(g))
print ''
if n_not_found > 0:
print ' not found: %d / %d = %f' % (n_not_found, n_not_found + n_found, n_not_found / float(n_not_found + n_found))
class IhhhmmmParser(object):
def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir, remove_N_nukes=True)
self.perfplotter = PerformancePlotter(self.germline_seqs, self.args.plotdir, 'ihhhmmm')
self.details = OrderedDict()
self.failtails = {}
self.n_partially_failed = 0
# get sequence info that was passed to ihhhmmm
self.siminfo = OrderedDict()
self.sim_need = [] # list of queries that we still need to find
with opener('r')(self.args.simfname) as seqfile:
reader = csv.DictReader(seqfile)
iline = 0
for line in reader:
if self.args.queries != None and line['unique_id'] not in self.args.queries:
continue
self.siminfo[line['unique_id']] = line
self.sim_need.append(line['unique_id'])
iline += 1
if args.n_max_queries > 0 and iline >= args.n_max_queries:
break
fostream_names = glob.glob(self.args.indir + '/*.fostream')
fostream_names.sort() # maybe already sorted?
for infname in fostream_names:
if len(self.sim_need) == 0:
break
# try to get whatever you can for the failures
unique_ids = self.find_partial_failures(infname) # returns list of unique ids in this file
with opener('r')(infname) as infile:
self.parse_file(infile, unique_ids)
# now check that we got results for all the queries we wanted
n_failed = 0
for unique_id in self.siminfo:
if unique_id not in self.details and unique_id not in self.failtails:
print '%-20s no info' % unique_id
self.perfplotter.add_fail()
n_failed += 1
print ''
print 'partially failed: %d / %d = %.2f' % (self.n_partially_failed, len(self.siminfo), float(self.n_partially_failed) / len(self.siminfo))
print 'failed: %d / %d = %.2f' % (n_failed, len(self.siminfo), float(n_failed) / len(self.siminfo))
print ''
self.perfplotter.plot()
# ----------------------------------------------------------------------------------------
def parse_file(self, infile, unique_ids):
fk = FileKeeper(infile.readlines())
i_id = 0
while not fk.eof and len(self.sim_need) > 0:
self.parse_detail(fk, unique_ids[i_id])
i_id += 1
# ----------------------------------------------------------------------------------------
def parse_detail(self, fk, unique_id):
assert fk.iline < len(fk.lines)
while fk.line[1] != 'Details':
fk.increment()
if fk.eof:
return
fk.increment()
info = {}
info['unique_id'] = unique_id
for begin_line, column, index, required, default in line_order:
if fk.line[0].find(begin_line) != 0:
if required:
print 'oop', begin_line, fk.line
sys.exit()
else:
info[column] = default
continue
if column != '':
info[column] = clean_value(column, fk.line[index])
# if '[' in info[column]:
# print 'added', column, clean_value(column, fk.line[index])
if column.find('_gene') == 1:
region = column[0]
info[region + '_5p_del'] = int(fk.line[fk.line.index('start:') + 1]) - 1 # NOTE their indices are 1-based
gl_length = int(fk.line[fk.line.index('gene:') + 1]) - 1
match_end = int(fk.line[fk.line.index('end:') + 1]) - 1
assert gl_length >= match_end
info[region + '_3p_del'] = gl_length - match_end
fk.increment()
if unique_id not in self.sim_need:
while not fk.eof and fk.line[1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
return
info['fv_insertion'] = ''
info['jf_insertion'] = ''
info['seq'] = info['v_qr_seq'] + info['vd_insertion'] + info['d_qr_seq'] + info['dj_insertion'] + info['j_qr_seq']
if '-' in info['seq']:
print 'ERROR found a dash in %s, returning failure' % unique_id
while not fk.eof and fk.line[1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
return
if info['seq'] not in self.siminfo[unique_id]['seq']: # arg. I can't do != because it tacks on v left and j right deletions
print 'ERROR didn\'t find the right sequence for %s' % unique_id
print ' ', info['seq']
print ' ', self.siminfo[unique_id]['seq']
sys.exit()
if self.args.debug:
print unique_id
utils.print_reco_event(self.germline_seqs, self.siminfo[unique_id], label='true:', extra_str=' ')
utils.print_reco_event(self.germline_seqs, info, label='inferred:', extra_str=' ')
for region in utils.regions:
if info[region + '_gene'] not in self.germline_seqs[region]:
print 'ERROR %s not in germlines' % info[region + '_gene']
assert False
gl_seq = info[region + '_gl_seq']
if '[' in gl_seq: # ambiguous
for nuke in utils.nukes:
gl_seq = gl_seq.replace('[', nuke)
if gl_seq in self.germline_seqs[region][info[region + '_gene']]:
print ' replaced [ with %s' % nuke
break
info[region + '_gl_seq'] = gl_seq
if info[region + '_gl_seq'] not in self.germline_seqs[region][info[region + '_gene']]:
print 'ERROR gl match not found for %s in %s' % (info[region + '_gene'], unique_id)
print ' ', info[region + '_gl_seq']
print ' ', self.germline_seqs[region][info[region + '_gene']]
self.perfplotter.add_partial_fail(self.siminfo[unique_id], info)
while not fk.eof and fk.line[1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
return
self.perfplotter.evaluate(self.siminfo[unique_id], info)
self.details[unique_id] = info
self.sim_need.remove(unique_id)
while not fk.eof and fk.line[1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
# ----------------------------------------------------------------------------------------
def find_partial_failures(self, fostream_name):
unique_ids = []
for line in open(fostream_name.replace('.fostream', '')).readlines():
if len(self.sim_need) == 0:
return
if len(line.strip()) == 0: # skip blank lines
continue
line = line.replace('"', '')
line = line.split(';')
unique_id = line[0]
if 'NA' not in line: # skip lines that were ok
unique_ids.append(unique_id)
continue
if unique_id not in self.sim_need:
continue
if unique_id not in self.siminfo:
continue # not looking for this <unique_id> a.t.m.
info = {}
info['unique_id'] = unique_id
for stuff in line:
for region in utils.regions: # add the first instance of IGH[VDJ] (if it's there at all)
if 'IGH'+region.upper() in stuff and region+'_gene' not in info:
genes = re.findall('IGH' + region.upper() + '[^ ][^ ]*', stuff)
if len(genes) == 0:
print 'ERROR no %s genes in %s' % (region, stuff)
gene = genes[0]
if gene not in self.germline_seqs[region]:
print 'ERROR bad gene %s for %s' % (gene, unique_id)
sys.exit()
info[region + '_gene'] = gene
self.perfplotter.add_partial_fail(self.siminfo[unique_id], info)
if self.args.debug:
print '%-20s partial fail %s %s %s' % (unique_id,
utils.color_gene(info['v_gene']) if 'v_gene' in info else '',
utils.color_gene(info['d_gene']) if 'd_gene' in info else '',
utils.color_gene(info['j_gene']) if 'j_gene' in info else ''),
print ' (true %s %s %s)' % tuple([self.siminfo[unique_id][region + '_gene'] for region in utils.regions])
self.failtails[unique_id] = info
self.n_partially_failed += 1
self.sim_need.remove(unique_id)
return unique_ids
def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir)
perfplotter = PerformancePlotter(self.germline_seqs, self.args.plotdir,
'imgt')
# get sequence info that was passed to imgt
self.seqinfo = {}
with opener('r')(self.args.simfname) as simfile:
reader = csv.DictReader(simfile)
iline = 0
for line in reader:
if self.args.queries != None and line[
'unique_id'] not in self.args.queries:
continue
if len(re.findall('_[FP]', line['j_gene'])) > 0:
line['j_gene'] = line['j_gene'].replace(
re.findall('_[FP]', line['j_gene'])[0], '')
self.seqinfo[line['unique_id']] = line
iline += 1
if self.args.n_queries > 0 and iline >= self.args.n_queries:
break
paragraphs, csv_info = None, None
if self.args.infname != None and '.html' in self.args.infname:
print 'reading', self.args.infname
with opener('r')(self.args.infname) as infile:
soup = BeautifulSoup(infile)
paragraphs = soup.find_all('pre')
summarydir = self.args.indir[:self.args.indir.rfind(
'/'
)] # one directoy up from <indir>, which has the detailed per-sequence files
summary_fname = glob.glob(summarydir + '/1_Summary_*.txt')
assert len(summary_fname) == 1
summary_fname = summary_fname[0]
get_genes_to_skip(summary_fname, self.germline_seqs)
n_failed, n_skipped, n_total, n_not_found, n_found = 0, 0, 0, 0, 0
for unique_id in self.seqinfo:
if self.args.debug:
print unique_id,
imgtinfo = []
# print 'true'
# utils.print_reco_event(self.germline_seqs, self.seqinfo[unique_id])
if self.args.infname != None and '.html' in self.args.infname:
for pre in paragraphs: # NOTE this loops over everything an awful lot of times. Shouldn't really matter for now, though
if unique_id in pre.text:
imgtinfo.append(pre.text)
else:
n_total += 1
assert self.args.infname == None
infnames = glob.glob(self.args.indir + '/' + unique_id + '*')
assert len(infnames) <= 1
if len(infnames) != 1:
if self.args.debug:
print ' couldn\'t find it'
n_not_found += 1
continue
n_found += 1
with opener('r')(infnames[0]) as infile:
full_text = infile.read()
if len(
re.findall('[123]. Alignment for [VDJ]-GENE',
full_text)) < 3:
failregions = re.findall(
'No [VDJ]-GENE has been identified', full_text)
if self.args.debug and len(failregions) > 0:
print ' ', failregions
n_failed += 1
continue
# loop over the paragraphs I want
position = full_text.find(unique_id) # don't need this one
for ir in range(4):
position = full_text.find(unique_id, position + 1)
pgraph = full_text[position:full_text.
find('\n\n', position + 1)]
if 'insertion(s) and/or deletion(s) which are not dealt in this release' in pgraph:
ir -= 1
continue
imgtinfo.append(pgraph) # query seq paragraph
if len(imgtinfo) == 0:
print '%s no info' % unique_id
continue
else:
if self.args.debug:
print ''
line = self.parse_query_text(unique_id, imgtinfo)
if 'skip_gene' in line:
# assert self.args.skip_missing_genes
n_skipped += 1
continue
try:
assert 'failed' not in line
joinparser.add_insertions(line, debug=self.args.debug)
joinparser.resolve_overlapping_matches(
line, debug=False, germlines=self.germline_seqs)
except (AssertionError, KeyError):
print ' giving up'
n_failed += 1
perfplotter.add_partial_fail(self.seqinfo[unique_id], line)
# print ' perfplotter: not sure what to do with a fail'
continue
perfplotter.evaluate(self.seqinfo[unique_id], line)
if self.args.debug:
utils.print_reco_event(self.germline_seqs,
self.seqinfo[unique_id],
label='true:')
utils.print_reco_event(self.germline_seqs,
line,
label='inferred:')
perfplotter.plot()
print 'failed: %d / %d = %f' % (n_failed, n_total,
float(n_failed) / n_total)
print 'skipped: %d / %d = %f' % (n_skipped, n_total,
float(n_skipped) / n_total)
print ' ',
for g, n in genes_actually_skipped.items():
print ' %d %s' % (n, utils.color_gene(g))
print ''
if n_not_found > 0:
print ' not found: %d / %d = %f' % (n_not_found, n_not_found +
n_found, n_not_found /
float(n_not_found + n_found))
class IhhhmmmParser(object):
def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir, remove_N_nukes=True)
self.perfplotter = PerformancePlotter(self.germline_seqs, self.args.plotdir, "ihhhmmm")
self.details = OrderedDict()
self.failtails = {}
self.n_partially_failed = 0
# get sequence info that was passed to ihhhmmm
self.siminfo = OrderedDict()
self.sim_need = [] # list of queries that we still need to find
with opener("r")(self.args.simfname) as seqfile:
reader = csv.DictReader(seqfile)
iline = 0
for line in reader:
if self.args.queries != None and line["unique_id"] not in self.args.queries:
continue
self.siminfo[line["unique_id"]] = line
self.sim_need.append(line["unique_id"])
iline += 1
if args.n_queries > 0 and iline >= args.n_queries:
break
fostream_names = glob.glob(self.args.indir + "/*.fostream")
if len(fostream_names) == 0:
raise Exception("no fostreams found in %s" % args.indir)
fostream_names.sort() # maybe already sorted?
for infname in fostream_names:
if len(self.sim_need) == 0:
break
# try to get whatever you can for the failures
unique_ids = self.find_partial_failures(infname) # returns list of unique ids in this file
with opener("r")(infname) as infile:
self.parse_file(infile, unique_ids)
# now check that we got results for all the queries we wanted
n_failed = 0
for unique_id in self.siminfo:
if unique_id not in self.details and unique_id not in self.failtails:
print "%-20s no info" % unique_id
self.perfplotter.add_fail()
n_failed += 1
print ""
print "partially failed: %d / %d = %.2f" % (
self.n_partially_failed,
len(self.siminfo),
float(self.n_partially_failed) / len(self.siminfo),
)
print "failed: %d / %d = %.2f" % (n_failed, len(self.siminfo), float(n_failed) / len(self.siminfo))
print ""
self.perfplotter.plot()
# ----------------------------------------------------------------------------------------
def parse_file(self, infile, unique_ids):
fk = FileKeeper(infile.readlines())
i_id = 0
while not fk.eof and len(self.sim_need) > 0:
self.parse_detail(fk, unique_ids[i_id])
i_id += 1
# ----------------------------------------------------------------------------------------
def parse_detail(self, fk, unique_id):
assert fk.iline < len(fk.lines)
while fk.line[1] != "Details":
fk.increment()
if fk.eof:
return
fk.increment()
info = {}
info["unique_id"] = unique_id
for begin_line, column, index, required, default in line_order:
if fk.line[0].find(begin_line) != 0:
if required:
print "oop", begin_line, fk.line
sys.exit()
else:
info[column] = default
continue
if column != "":
info[column] = clean_value(column, fk.line[index])
# if '[' in info[column]:
# print 'added', column, clean_value(column, fk.line[index])
if column.find("_gene") == 1:
region = column[0]
info[region + "_5p_del"] = (
int(fk.line[fk.line.index("start:") + 1]) - 1
) # NOTE their indices are 1-based
gl_length = int(fk.line[fk.line.index("gene:") + 1]) - 1
match_end = int(fk.line[fk.line.index("end:") + 1]) - 1
assert gl_length >= match_end
info[region + "_3p_del"] = gl_length - match_end
fk.increment()
if unique_id not in self.sim_need:
while not fk.eof and fk.line[1] != "Details": # skip stuff until start of next Detail block
fk.increment()
return
info["fv_insertion"] = ""
info["jf_insertion"] = ""
info["seq"] = (
info["v_qr_seq"] + info["vd_insertion"] + info["d_qr_seq"] + info["dj_insertion"] + info["j_qr_seq"]
)
if "-" in info["seq"]:
print "ERROR found a dash in %s, returning failure" % unique_id
while not fk.eof and fk.line[1] != "Details": # skip stuff until start of next Detail block
fk.increment()
return
if (
info["seq"] not in self.siminfo[unique_id]["seq"]
): # arg. I can't do != because it tacks on v left and j right deletions
print "ERROR didn't find the right sequence for %s" % unique_id
print " ", info["seq"]
print " ", self.siminfo[unique_id]["seq"]
sys.exit()
if self.args.debug:
print unique_id
for region in utils.regions:
infer_gene = info[region + "_gene"]
true_gene = self.siminfo[unique_id][region + "_gene"]
if utils.are_alleles(infer_gene, true_gene):
regionstr = utils.color("bold", utils.color("blue", region))
truestr = "" #'(originally %s)' % match_name
else:
regionstr = utils.color("bold", utils.color("red", region))
truestr = "(true: %s)" % utils.color_gene(true_gene).replace(region, "")
print " %s %s %s" % (regionstr, utils.color_gene(infer_gene).replace(region, ""), truestr)
utils.print_reco_event(self.germline_seqs, self.siminfo[unique_id], label="true:", extra_str=" ")
utils.print_reco_event(self.germline_seqs, info, label="inferred:", extra_str=" ")
for region in utils.regions:
if info[region + "_gene"] not in self.germline_seqs[region]:
print "ERROR %s not in germlines" % info[region + "_gene"]
assert False
gl_seq = info[region + "_gl_seq"]
if "[" in gl_seq: # ambiguous
for nuke in utils.nukes:
gl_seq = gl_seq.replace("[", nuke)
if gl_seq in self.germline_seqs[region][info[region + "_gene"]]:
print " replaced [ with %s" % nuke
break
info[region + "_gl_seq"] = gl_seq
if info[region + "_gl_seq"] not in self.germline_seqs[region][info[region + "_gene"]]:
print "ERROR gl match not found for %s in %s" % (info[region + "_gene"], unique_id)
print " ", info[region + "_gl_seq"]
print " ", self.germline_seqs[region][info[region + "_gene"]]
self.perfplotter.add_partial_fail(self.siminfo[unique_id], info)
while not fk.eof and fk.line[1] != "Details": # skip stuff until start of next Detail block
fk.increment()
return
self.perfplotter.evaluate(self.siminfo[unique_id], info)
self.details[unique_id] = info
self.sim_need.remove(unique_id)
while not fk.eof and fk.line[1] != "Details": # skip stuff until start of next Detail block
fk.increment()
# ----------------------------------------------------------------------------------------
def find_partial_failures(self, fostream_name):
unique_ids = []
for line in open(fostream_name.replace(".fostream", "")).readlines():
if len(self.sim_need) == 0:
return
if len(line.strip()) == 0: # skip blank lines
continue
line = line.replace('"', "")
line = line.split(";")
unique_id = line[0]
if "NA" not in line: # skip lines that were ok
unique_ids.append(unique_id)
continue
if unique_id not in self.sim_need:
continue
if unique_id not in self.siminfo:
continue # not looking for this <unique_id> a.t.m.
info = {}
info["unique_id"] = unique_id
for stuff in line:
for region in utils.regions: # add the first instance of IGH[VDJ] (if it's there at all)
if "IGH" + region.upper() in stuff and region + "_gene" not in info:
genes = re.findall("IGH" + region.upper() + "[^ ][^ ]*", stuff)
if len(genes) == 0:
print "ERROR no %s genes in %s" % (region, stuff)
gene = genes[0]
if gene not in self.germline_seqs[region]:
print "ERROR bad gene %s for %s" % (gene, unique_id)
sys.exit()
info[region + "_gene"] = gene
self.perfplotter.add_partial_fail(self.siminfo[unique_id], info)
if self.args.debug:
print "%-20s partial fail %s %s %s" % (
unique_id,
utils.color_gene(info["v_gene"]) if "v_gene" in info else "",
utils.color_gene(info["d_gene"]) if "d_gene" in info else "",
utils.color_gene(info["j_gene"]) if "j_gene" in info else "",
),
print " (true %s %s %s)" % tuple(
[self.siminfo[unique_id][region + "_gene"] for region in utils.regions]
)
self.failtails[unique_id] = info
self.n_partially_failed += 1
self.sim_need.remove(unique_id)
return unique_ids
class IgblastParser(object):
def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir,
remove_N_nukes=True)
self.perfplotter = PerformancePlotter(self.germline_seqs,
self.args.plotdir, 'igblast')
self.n_total, self.n_partially_failed = 0, 0
# get sequence info that was passed to igblast
self.seqinfo = {}
with opener('r')(self.args.simfname) as simfile:
reader = csv.DictReader(simfile)
iline = 0
for line in reader:
if self.args.n_max_queries > 0 and iline >= self.args.n_max_queries:
break
iline += 1
if self.args.queries != None and int(
line['unique_id']) not in self.args.queries:
continue
if len(re.findall('_[FP]', line['j_gene'])) > 0:
line['j_gene'] = line['j_gene'].replace(
re.findall('_[FP]', line['j_gene'])[0], '')
self.seqinfo[int(line['unique_id'])] = line
paragraphs = None
print 'reading', self.args.infname
info = {}
with opener('r')(self.args.infname) as infile:
line = infile.readline()
# first find the start of the next query's section
while line.find('<b>Query=') != 0:
line = infile.readline()
# then keep going till eof
iquery = 0
while line != '':
if self.args.n_max_queries > 0 and iquery >= self.args.n_max_queries:
break
# first find the query name
query_name = int(line.split()[1])
# and collect the lines for this query
query_lines = []
line = infile.readline()
while line.find('<b>Query=') != 0:
query_lines.append(line.strip())
line = infile.readline()
if line == '':
break
iquery += 1
# then see if we want this query
if self.args.queries != None and query_name not in self.args.queries:
continue
if query_name not in self.seqinfo:
print 'ERROR %d not in reco info' % query_name
sys.exit()
if self.args.debug:
print query_name
# and finally add the query to <info[query_name]>
info[query_name] = {'unique_id': query_name}
self.n_total += 1
self.process_query(info[query_name], query_name, query_lines)
self.perfplotter.plot()
print 'partially failed: %d / %d = %f' % (
self.n_partially_failed, self.n_total,
float(self.n_partially_failed) / self.n_total)
# ----------------------------------------------------------------------------------------
def process_query(self, qr_info, query_name, query_lines):
# split query_lines up into blocks
blocks = []
for line in query_lines:
if line.find('Query_') == 0:
blocks.append([])
if len(line) == 0:
continue
if len(re.findall('<a name=#_[0-9][0-9]*_IGH',
line)) == 0 and line.find('Query_') != 0:
continue
if len(blocks) == 0:
print 'wtf? %s' % query_name # it's probably kicking a reverse match
self.perfplotter.add_partial_fail(
self.seqinfo[query_name],
qr_info) # NOTE that's really a total failure
self.n_partially_failed += 1
return
blocks[-1].append(line)
# then process each block
for block in blocks:
self.process_single_block(block, query_name, qr_info)
if 'fail' in qr_info:
self.perfplotter.add_partial_fail(self.seqinfo[query_name],
qr_info)
self.n_partially_failed += 1
return
for region in utils.regions:
if region + '_gene' not in qr_info:
print ' ERROR no %s match for %d' % (region, query_name)
self.perfplotter.add_partial_fail(self.seqinfo[query_name],
qr_info)
self.n_partially_failed += 1
return
# expand v match to left end and j match to right end
qr_info['v_5p_del'] = 0
qr_info['fv_insertion'] = ''
if qr_info['match_start'] > 0:
if self.args.debug:
print ' add to v left:', self.seqinfo[query_name][
'seq'][:qr_info['match_start']]
qr_info['seq'] = self.seqinfo[query_name][
'seq'][:qr_info['match_start']] + qr_info['seq']
qr_info['j_3p_del'] = 0
qr_info['jf_insertion'] = ''
if len(self.seqinfo[query_name]['seq']) > qr_info['match_end']:
if self.args.debug:
print ' add to j right:', self.seqinfo[query_name][
'seq'][qr_info['match_end'] -
len(self.seqinfo[query_name]['seq']):]
qr_info['seq'] = qr_info['seq'] + self.seqinfo[query_name]['seq'][
qr_info['match_end'] - len(self.seqinfo[query_name]['seq']):]
for boundary in utils.boundaries:
start = qr_info[boundary[0] + '_qr_bounds'][1]
end = qr_info[boundary[1] + '_qr_bounds'][0]
qr_info[boundary + '_insertion'] = qr_info['seq'][start:end]
for region in utils.regions:
start = qr_info[region + '_qr_bounds'][0]
end = qr_info[region + '_qr_bounds'][1]
qr_info[region + '_qr_seq'] = qr_info['seq'][start:end]
try:
resolve_overlapping_matches(qr_info, self.args.debug,
self.germline_seqs)
except AssertionError:
print 'ERROR apportionment failed on %s' % query_name
self.perfplotter.add_partial_fail(self.seqinfo[query_name],
qr_info)
self.n_partially_failed += 1
return
if self.args.debug:
print ' query seq:', qr_info['seq']
for region in utils.regions:
print ' %s %3d %3d %s %s' % (
region, qr_info[region + '_qr_bounds'][0],
qr_info[region + '_qr_bounds'][1],
utils.color_gene(qr_info[region + '_gene']),
qr_info[region + '_gl_seq'])
for boundary in utils.boundaries:
start = qr_info[boundary[0] + '_qr_bounds'][1]
end = qr_info[boundary[1] + '_qr_bounds'][0]
qr_info[boundary + '_insertion'] = qr_info['seq'][start:end]
if self.args.debug:
print ' ', boundary, qr_info[boundary + '_insertion']
self.perfplotter.evaluate(self.seqinfo[query_name], qr_info)
# for key, val in qr_info.items():
# print key, val
if self.args.debug:
utils.print_reco_event(self.germline_seqs,
self.seqinfo[query_name],
label='true:',
extra_str=' ')
utils.print_reco_event(self.germline_seqs, qr_info, extra_str=' ')
# ----------------------------------------------------------------------------------------
def process_single_block(self, block, query_name, qr_info):
assert block[0].find('Query_') == 0
vals = block[0].split()
qr_start = int(
vals[1]) - 1 # converting from one-indexed to zero-indexed
qr_seq = vals[2]
qr_end = int(
vals[3]
) # ...and from inclusive of both bounds to normal programming conventions
if qr_seq not in self.seqinfo[query_name]['seq']:
if '-' in qr_seq:
print ' WARNING insertion inside query seq for %s, treating as partial failure' % query_name
qr_info['fail'] = True
return
else:
print ' ERROR query seq from igblast info not found in original query seq for %d' % query_name
print ' %s' % qr_seq
print ' %s' % self.seqinfo[query_name]['seq']
sys.exit()
if 'seq' in qr_info:
qr_info['seq'] += qr_seq
else:
qr_info['seq'] = qr_seq
# keep track of the absolute first and absolute last bases matched so we can later work out the fv and jf insertions
if 'match_start' not in qr_info or qr_start < qr_info['match_start']:
qr_info['match_start'] = qr_start
if 'match_end' not in qr_info or qr_end > qr_info['match_end']:
qr_info['match_end'] = qr_end
if self.args.debug:
print ' query: %3d %3d %s' % (qr_start, qr_end, qr_seq)
for line in block[1:]:
gene = line[line.rfind('IGH'):line.rfind('</a>')]
region = utils.get_region(gene)
if gene not in self.germline_seqs[region]:
print ' ERROR %s not found in germlines' % gene
qr_info['fail'] = True
return
vals = line.split()
gl_start = int(
vals[-3]) - 1 # converting from one-indexed to zero-indexed
gl_seq = vals[-2]
gl_end = int(
vals[-1]
) # ...and from inclusive of both bounds to normal programming conventions
if region + '_gene' in qr_info:
if qr_info[region + '_gene'] == gene:
if self.args.debug:
print ' %s match: %s' % (
region, clean_alignment_crap(qr_seq, gl_seq))
qr_info[region + '_gl_seq'] = qr_info[
region + '_gl_seq'] + clean_alignment_crap(
qr_seq, gl_seq)
assert gl_end <= len(self.germline_seqs[region][gene])
qr_info[region + '_3p_del'] = len(
self.germline_seqs[region][gene]) - gl_end
qr_info[region +
'_qr_bounds'] = (qr_info[region + '_qr_bounds'][0],
find_qr_bounds(
qr_start, qr_end, gl_seq)[1])
else:
continue
else:
qr_info[region + '_gene'] = gene
qr_info[region + '_gl_seq'] = clean_alignment_crap(
qr_seq, gl_seq)
# deletions
qr_info[region + '_5p_del'] = gl_start
assert gl_end <= len(self.germline_seqs[region][gene])
qr_info[region + '_3p_del'] = len(
self.germline_seqs[region][gene]) - gl_end
# bounds
qr_info[region + '_qr_bounds'] = find_qr_bounds(
qr_start, qr_end, gl_seq)
if self.args.debug:
print ' %s match: %s' % (
region, clean_alignment_crap(qr_seq, gl_seq))
class IgblastParser(object):
def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir, remove_N_nukes=True)
self.perfplotter = PerformancePlotter(self.germline_seqs, self.args.plotdir, 'igblast')
self.n_total, self.n_partially_failed, self.n_skipped = 0, 0, 0
# get sequence info that was passed to igblast
self.seqinfo = {}
with opener('r')(self.args.simfname) as simfile:
reader = csv.DictReader(simfile)
iline = 0
for line in reader:
if self.args.n_queries > 0 and iline >= self.args.n_queries:
break
iline += 1
if self.args.queries != None and int(line['unique_id']) not in self.args.queries:
continue
if len(re.findall('_[FP]', line['j_gene'])) > 0:
line['j_gene'] = line['j_gene'].replace(re.findall('_[FP]', line['j_gene'])[0], '')
self.seqinfo[int(line['unique_id'])] = line
print 'reading', self.args.infname
get_genes_to_skip(self.args.infname, self.germline_seqs, method='igblast', debug=False)
paragraphs = None
info = {}
with opener('r')(self.args.infname) as infile:
line = infile.readline()
# first find the start of the next query's section
while line.find('<b>Query=') != 0:
line = infile.readline()
# then keep going till eof
iquery = 0
while line != '':
if self.args.n_queries > 0 and iquery >= self.args.n_queries:
break
# first find the query name
query_name = int(line.split()[1])
# and collect the lines for this query
query_lines = []
line = infile.readline()
while line.find('<b>Query=') != 0:
query_lines.append(line.strip())
line = infile.readline()
if line == '':
break
iquery += 1
# then see if we want this query
if self.args.queries != None and query_name not in self.args.queries:
continue
if query_name not in self.seqinfo:
print 'ERROR %d not in reco info' % query_name
sys.exit()
if self.args.debug:
print query_name
# and finally add the query to <info[query_name]>
info[query_name] = {'unique_id':query_name}
self.n_total += 1
self.process_query(info[query_name], query_name, query_lines)
self.perfplotter.plot()
print 'partially failed: %d / %d = %f' % (self.n_partially_failed, self.n_total, float(self.n_partially_failed) / self.n_total)
print 'skipped: %d / %d = %f' % (self.n_skipped, self.n_total, float(self.n_skipped) / self.n_total)
for g, n in genes_actually_skipped.items():
print ' %d %s' % (n, utils.color_gene(g))
# ----------------------------------------------------------------------------------------
def process_query(self, qr_info, query_name, query_lines):
# split query_lines up into blocks
blocks = []
for line in query_lines:
if line.find('Query_') == 0:
blocks.append([])
if len(line) == 0:
continue
if len(re.findall('<a name=#_[0-9][0-9]*_IGH', line)) == 0 and line.find('Query_') != 0:
continue
if len(blocks) == 0:
print 'wtf? %s' % query_name # it's probably kicking a reverse match
self.perfplotter.add_partial_fail(self.seqinfo[query_name], qr_info) # NOTE that's really a total failure
self.n_partially_failed += 1
return
blocks[-1].append(line)
# then process each block
for block in blocks:
self.process_single_block(block, query_name, qr_info)
if 'skip_gene' in qr_info:
self.n_skipped += 1
return
if 'fail' in qr_info:
self.perfplotter.add_partial_fail(self.seqinfo[query_name], qr_info)
self.n_partially_failed += 1
return
for region in utils.regions:
if region + '_gene' not in qr_info:
print ' %d: no %s match' % (query_name, region)
self.perfplotter.add_partial_fail(self.seqinfo[query_name], qr_info)
self.n_partially_failed += 1
return
# expand v match to left end and j match to right end
qr_info['v_5p_del'] = 0
qr_info['fv_insertion'] = ''
if qr_info['match_start'] > 0:
if self.args.debug:
print ' add to v left:', self.seqinfo[query_name]['seq'][ : qr_info['match_start']]
qr_info['seq'] = self.seqinfo[query_name]['seq'][ : qr_info['match_start']] + qr_info['seq']
qr_info['j_3p_del'] = 0
qr_info['jf_insertion'] = ''
if len(self.seqinfo[query_name]['seq']) > qr_info['match_end']:
if self.args.debug:
print ' add to j right:', self.seqinfo[query_name]['seq'][ qr_info['match_end'] - len(self.seqinfo[query_name]['seq']) : ]
qr_info['seq'] = qr_info['seq'] + self.seqinfo[query_name]['seq'][ qr_info['match_end'] - len(self.seqinfo[query_name]['seq']) : ]
for boundary in utils.boundaries:
start = qr_info[boundary[0] + '_qr_bounds'][1]
end = qr_info[boundary[1] + '_qr_bounds'][0]
qr_info[boundary + '_insertion'] = qr_info['seq'][start : end]
for region in utils.regions:
start = qr_info[region + '_qr_bounds'][0]
end = qr_info[region + '_qr_bounds'][1]
qr_info[region + '_qr_seq'] = qr_info['seq'][start : end]
try:
resolve_overlapping_matches(qr_info, self.args.debug, self.germline_seqs)
except AssertionError:
print ' %s: apportionment failed' % query_name
self.perfplotter.add_partial_fail(self.seqinfo[query_name], qr_info)
self.n_partially_failed += 1
return
if self.args.debug:
print ' query seq:', qr_info['seq']
for region in utils.regions:
true_gene = self.seqinfo[query_name][region + '_gene']
infer_gene = qr_info[region + '_gene']
if utils.are_alleles(infer_gene, true_gene):
regionstr = utils.color('bold', utils.color('blue', region))
truestr = '' #'(originally %s)' % match_name
else:
regionstr = utils.color('bold', utils.color('red', region))
truestr = '(true: %s)' % utils.color_gene(true_gene).replace(region, '')
# print ' %s %s %s' % (regionstr, utils.color_gene(infer_gene).replace(region, ''), truestr)
print ' %s %3d %3d %s %s %s' % (regionstr, qr_info[region + '_qr_bounds'][0], qr_info[region + '_qr_bounds'][1], utils.color_gene(infer_gene).replace(region, ''), truestr, qr_info[region + '_gl_seq'])
for boundary in utils.boundaries:
start = qr_info[boundary[0] + '_qr_bounds'][1]
end = qr_info[boundary[1] + '_qr_bounds'][0]
qr_info[boundary + '_insertion'] = qr_info['seq'][start : end]
if self.args.debug:
print ' ', boundary, qr_info[boundary + '_insertion']
self.perfplotter.evaluate(self.seqinfo[query_name], qr_info)
# for key, val in qr_info.items():
# print key, val
if self.args.debug:
utils.print_reco_event(self.germline_seqs, self.seqinfo[query_name], label='true:', extra_str=' ')
utils.print_reco_event(self.germline_seqs, qr_info, extra_str=' ')
# ----------------------------------------------------------------------------------------
def process_single_block(self, block, query_name, qr_info):
assert block[0].find('Query_') == 0
vals = block[0].split()
qr_start = int(vals[1]) - 1 # converting from one-indexed to zero-indexed
qr_seq = vals[2]
qr_end = int(vals[3]) # ...and from inclusive of both bounds to normal programming conventions
if qr_seq not in self.seqinfo[query_name]['seq']:
if '-' in qr_seq:
print ' %s: insertion inside query seq, treating as partial failure' % query_name
qr_info['fail'] = True
return
else:
print ' ERROR query seq from igblast info not found in original query seq for %d' % query_name
print ' %s' % qr_seq
print ' %s' % self.seqinfo[query_name]['seq']
sys.exit()
if 'seq' in qr_info:
qr_info['seq'] += qr_seq
else:
qr_info['seq'] = qr_seq
# keep track of the absolute first and absolute last bases matched so we can later work out the fv and jf insertions
if 'match_start' not in qr_info or qr_start < qr_info['match_start']:
qr_info['match_start'] = qr_start
if 'match_end' not in qr_info or qr_end > qr_info['match_end']:
qr_info['match_end'] = qr_end
# ----------------------------------------------------------------------------------------
# skipping bullshit
def skip_gene(gene):
if self.args.debug:
print ' %s in list of genes to skip' % utils.color_gene(gene)
if gene not in genes_actually_skipped:
genes_actually_skipped[gene] = 0
genes_actually_skipped[gene] += 1
qr_info['skip_gene'] = True
if self.args.debug:
print ' query: %3d %3d %s' % (qr_start, qr_end, qr_seq)
for line in block[1:]:
gene = line[line.rfind('IGH') : line.rfind('</a>')]
region = utils.get_region(gene)
true_gene = self.seqinfo[query_name][region + '_gene']
for gset in equivalent_genes:
if gene in gset and true_gene in gset and gene != true_gene: # if the true gene and the inferred gene are in the same equivalence set, treat it as correct, i.e. just pretend it inferred the right name
if self.args.debug:
print ' %s: replacing name %s with true name %s' % (query_name, gene, true_gene)
gene = true_gene
if gene in just_always_friggin_skip:
continue # go on to the next match
if not self.args.dont_skip_or15_genes and '/OR1' in true_gene:
skip_gene(true_gene)
return
if self.args.skip_missing_genes:
if gene in genes_to_skip:
continue # go on to the next match
# skip_gene(gene)
# return
if true_gene in genes_to_skip:
skip_gene(true_gene)
return
if gene not in self.germline_seqs[region]:
print ' %s: %s not in germlines (skipping)' % (query_name, gene)
skip_gene(gene)
return
vals = line.split()
gl_start = int(vals[-3]) - 1 # converting from one-indexed to zero-indexed
gl_seq = vals[-2]
gl_end = int(vals[-1]) # ...and from inclusive of both bounds to normal programming conventions
if region + '_gene' in qr_info:
if qr_info[region + '_gene'] == gene:
if self.args.debug:
print ' %s match: %s' % (region, clean_alignment_crap(qr_seq, gl_seq))
qr_info[region + '_gl_seq'] = qr_info[region + '_gl_seq'] + clean_alignment_crap(qr_seq, gl_seq)
if gl_end > len(self.germline_seqs[region][gene]): # not really sure what's wrong... but it seems to be rare
qr_info['fail'] = True
return
qr_info[region + '_3p_del'] = len(self.germline_seqs[region][gene]) - gl_end
qr_info[region + '_qr_bounds'] = (qr_info[region + '_qr_bounds'][0], find_qr_bounds(qr_start, qr_end, gl_seq)[1])
else:
continue
else:
qr_info[region + '_gene'] = gene
qr_info[region + '_gl_seq'] = clean_alignment_crap(qr_seq, gl_seq)
# deletions
qr_info[region + '_5p_del'] = gl_start
assert gl_end <= len(self.germline_seqs[region][gene])
qr_info[region + '_3p_del'] = len(self.germline_seqs[region][gene]) - gl_end
# bounds
qr_info[region + '_qr_bounds'] = find_qr_bounds(qr_start, qr_end, gl_seq)
if self.args.debug:
print ' %s match: %s' % (region, clean_alignment_crap(qr_seq, gl_seq))
class IhhhmmmParser(object):
def __init__(self, args):
self.args = args
self.germline_seqs = utils.read_germlines(self.args.datadir,
remove_N_nukes=True)
self.perfplotter = PerformancePlotter(self.germline_seqs,
self.args.plotdir, 'ihhhmmm')
self.details = OrderedDict()
self.failtails = {}
self.n_partially_failed = 0
# get sequence info that was passed to ihhhmmm
self.siminfo = OrderedDict()
self.sim_need = [] # list of queries that we still need to find
with opener('r')(self.args.simfname) as seqfile:
reader = csv.DictReader(seqfile)
iline = 0
for line in reader:
if self.args.queries != None and line[
'unique_id'] not in self.args.queries:
continue
self.siminfo[line['unique_id']] = line
self.sim_need.append(line['unique_id'])
iline += 1
if args.n_max_queries > 0 and iline >= args.n_max_queries:
break
fostream_names = glob.glob(self.args.indir + '/*.fostream')
fostream_names.sort() # maybe already sorted?
for infname in fostream_names:
if len(self.sim_need) == 0:
break
# try to get whatever you can for the failures
unique_ids = self.find_partial_failures(
infname) # returns list of unique ids in this file
with opener('r')(infname) as infile:
self.parse_file(infile, unique_ids)
# now check that we got results for all the queries we wanted
n_failed = 0
for unique_id in self.siminfo:
if unique_id not in self.details and unique_id not in self.failtails:
print '%-20s no info' % unique_id
self.perfplotter.add_fail()
n_failed += 1
print ''
print 'partially failed: %d / %d = %.2f' % (
self.n_partially_failed, len(self.siminfo),
float(self.n_partially_failed) / len(self.siminfo))
print 'failed: %d / %d = %.2f' % (n_failed, len(
self.siminfo), float(n_failed) / len(self.siminfo))
print ''
self.perfplotter.plot()
# ----------------------------------------------------------------------------------------
def parse_file(self, infile, unique_ids):
fk = FileKeeper(infile.readlines())
i_id = 0
while not fk.eof and len(self.sim_need) > 0:
self.parse_detail(fk, unique_ids[i_id])
i_id += 1
# ----------------------------------------------------------------------------------------
def parse_detail(self, fk, unique_id):
assert fk.iline < len(fk.lines)
while fk.line[1] != 'Details':
fk.increment()
if fk.eof:
return
fk.increment()
info = {}
info['unique_id'] = unique_id
for begin_line, column, index, required, default in line_order:
if fk.line[0].find(begin_line) != 0:
if required:
print 'oop', begin_line, fk.line
sys.exit()
else:
info[column] = default
continue
if column != '':
info[column] = clean_value(column, fk.line[index])
# if '[' in info[column]:
# print 'added', column, clean_value(column, fk.line[index])
if column.find('_gene') == 1:
region = column[0]
info[region + '_5p_del'] = int(
fk.line[fk.line.index('start:') +
1]) - 1 # NOTE their indices are 1-based
gl_length = int(fk.line[fk.line.index('gene:') + 1]) - 1
match_end = int(fk.line[fk.line.index('end:') + 1]) - 1
assert gl_length >= match_end
info[region + '_3p_del'] = gl_length - match_end
fk.increment()
if unique_id not in self.sim_need:
while not fk.eof and fk.line[
1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
return
info['fv_insertion'] = ''
info['jf_insertion'] = ''
info['seq'] = info['v_qr_seq'] + info['vd_insertion'] + info[
'd_qr_seq'] + info['dj_insertion'] + info['j_qr_seq']
if '-' in info['seq']:
print 'ERROR found a dash in %s, returning failure' % unique_id
while not fk.eof and fk.line[
1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
return
if info['seq'] not in self.siminfo[unique_id][
'seq']: # arg. I can't do != because it tacks on v left and j right deletions
print 'ERROR didn\'t find the right sequence for %s' % unique_id
print ' ', info['seq']
print ' ', self.siminfo[unique_id]['seq']
sys.exit()
if self.args.debug:
print unique_id
utils.print_reco_event(self.germline_seqs,
self.siminfo[unique_id],
label='true:',
extra_str=' ')
utils.print_reco_event(self.germline_seqs,
info,
label='inferred:',
extra_str=' ')
for region in utils.regions:
if info[region + '_gene'] not in self.germline_seqs[region]:
print 'ERROR %s not in germlines' % info[region + '_gene']
assert False
gl_seq = info[region + '_gl_seq']
if '[' in gl_seq: # ambiguous
for nuke in utils.nukes:
gl_seq = gl_seq.replace('[', nuke)
if gl_seq in self.germline_seqs[region][info[region +
'_gene']]:
print ' replaced [ with %s' % nuke
break
info[region + '_gl_seq'] = gl_seq
if info[region + '_gl_seq'] not in self.germline_seqs[region][info[
region + '_gene']]:
print 'ERROR gl match not found for %s in %s' % (
info[region + '_gene'], unique_id)
print ' ', info[region + '_gl_seq']
print ' ', self.germline_seqs[region][info[region + '_gene']]
self.perfplotter.add_partial_fail(self.siminfo[unique_id],
info)
while not fk.eof and fk.line[
1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
return
self.perfplotter.evaluate(self.siminfo[unique_id], info)
self.details[unique_id] = info
self.sim_need.remove(unique_id)
while not fk.eof and fk.line[
1] != 'Details': # skip stuff until start of next Detail block
fk.increment()
# ----------------------------------------------------------------------------------------
def find_partial_failures(self, fostream_name):
unique_ids = []
for line in open(fostream_name.replace('.fostream', '')).readlines():
if len(self.sim_need) == 0:
return
if len(line.strip()) == 0: # skip blank lines
continue
line = line.replace('"', '')
line = line.split(';')
unique_id = line[0]
if 'NA' not in line: # skip lines that were ok
unique_ids.append(unique_id)
continue
if unique_id not in self.sim_need:
continue
if unique_id not in self.siminfo:
continue # not looking for this <unique_id> a.t.m.
info = {}
info['unique_id'] = unique_id
for stuff in line:
for region in utils.regions: # add the first instance of IGH[VDJ] (if it's there at all)
if 'IGH' + region.upper(
) in stuff and region + '_gene' not in info:
genes = re.findall(
'IGH' + region.upper() + '[^ ][^ ]*', stuff)
if len(genes) == 0:
print 'ERROR no %s genes in %s' % (region, stuff)
gene = genes[0]
if gene not in self.germline_seqs[region]:
print 'ERROR bad gene %s for %s' % (gene,
unique_id)
sys.exit()
info[region + '_gene'] = gene
self.perfplotter.add_partial_fail(self.siminfo[unique_id], info)
if self.args.debug:
print '%-20s partial fail %s %s %s' % (
unique_id, utils.color_gene(info['v_gene']) if 'v_gene'
in info else '', utils.color_gene(info['d_gene']) if
'd_gene' in info else '', utils.color_gene(info['j_gene'])
if 'j_gene' in info else ''),
print ' (true %s %s %s)' % tuple([
self.siminfo[unique_id][region + '_gene']
for region in utils.regions
])
self.failtails[unique_id] = info
self.n_partially_failed += 1
self.sim_need.remove(unique_id)
return unique_ids
