def Vcf(fh, sv, variant_id):
errFH(fh)
vcf_fh = None
if fh.endswith('.gz'):
import gzip
vcf_fh = gzip.open(fh, 'r')
else:
vcf_fh = open(fh, 'r')
with vcf_fh as f:
for l in f:
if l.startswith('#'): continue
s, e, r = -9, -9, tokenize(l)
if r != 0:
c, s = str(r[0]), int(r[1])
s -= 1
if len(r) < 8:
sys.stderr.write(
'WARNING: Skipping bad entry {} in VCF file {}. INFO column is required: {}\n'
.format(l.rstrip(), fh, __help_url__))
else:
for i in r[7].split(';'):
if 'SVTYPE=' in i: cl = i.replace('SVTYPE=', '')
if i.startswith('END=') and 'CI' not in i:
e = int(i.replace('END=', ''))
if s == -9 or e == -9:
sys.stderr.write(
'WARNING: Skipping bad entry {} in VCF file {}. SV entry missing positions: {}\n'
.format(l.rstrip(), fh, __help_url__))
else:
if checkCall(cl, l, fh) == True:
sv.append((c, s, e, cl))
variant_id[(str(c), int(s), int(e),
str(cl))] = str(r[2])
vcf_fh.close()
def load_clf(self, jsonfh=None):
import shutil
errFH(jsonfh)
clfs = {}
with open(jsonfh) as f:
clfs = json.load(f)
realpaths = []
for name in clfs:
print 'loading {} classifier ...'.format(name)
clf_dir = get_path() + '/resources/training_sets/' + name + '/'
if not os.path.exists(clf_dir): os.makedirs(clf_dir)
for x in clfs[name]:
clffh = str(clfs[name][x])
if not os.path.isfile(clffh):
sys.stderr.write(
'WARNING: {} does not exist. Please check the paths in {}\n'
.format(clffh, jsonfh))
else:
clfname = clffh.split('/').pop()
newpath = clf_dir + clfname
clfreplace = True
if os.path.isfile(newpath):
clfreplace = query_yes_no(
'WARNING: {} exists... Replace?'.format(newpath),
'no')
if clfreplace == True:
if self.clfs.get(name) == None: self.clfs[name] = {}
self.clfs[name][x] = newpath
realpaths.append(newpath)
shutil.copyfile(clffh, newpath)
for x in realpaths:
print 'installed classifier {}'.format(x)
print 'appending to {}... DO NOT ALTER {}'.format(self.json, self.json)
dump_json(self.json, self.clfs)
def __init__(self, peds=None):
self.sex = {}
self.males = {}
self.ids = []
for fh in peds:
errFH(fh)
with open(fh, 'r') as f:
for l in f:
r = tokenize(l)
if r != 0:
if len(r) < 5:
sys.stderr.write(
'WARNING {} does not contain 5 elements. PED files are formatted as:\nFamily ID Individual ID Paternal ID Maternal ID Sex(1=male;2=female;)\n'
.format(l))
else:
sample_id, sex = str(r[1]), int(r[4])
if sex != 1 and sex != 2:
sys.stderr.write(
'WARNING {} is not an accepted sex entry. Accepted entries: 1=male; 2=female. Error found here:{}\n'
.format(sex, l))
else:
self.sex[sample_id] = sex
self.ids.append(sample_id)
if sex == 1: self.males[sample_id] = 1
self.ids = sorted(list(set(self.ids)), key=str.lower)
def Bed(fh, sv):
errFH(fh)
with open(fh, 'r') as f:
for l in f:
if l.startswith('#'): continue
r = tokenize(l)
if r == 0 or len(r) < 4:
sys.stderr.write(
'WARNING: Skipping bad entry {} in BED file {}. BED format requirements: {}\n'
.format(l.rstrip(), fh, __help_url__))
elif r != 0 and checkCall(str(r[3]), l, fh) != False:
sv.append((str(r[0]), int(r[1]), int(r[2]), str(r[3])))
def bam_init(args=None, Ped=None, Snv=None, gen=None):
bams = []
for f in args:
errFH(f)
try:
pysam.AlignmentFile(f).check_index()
except ValueError:
sys.stderr.write(
'WARNING: {} is not indexed with samtools index. Skipping ...\n'
.format(f))
continue
except AttributeError:
sys.stderr.write(
'WARNING: {} appears to be in SAM format. Convert to BAM with `samtools view -bh {}` and index with samtools index\n'
.format(f, f))
continue
bam = Bam(f, Ped, gen)
if len(Snv) < 1:
print 'FATAL ERROR: SNV file(s) were not formatted correctly. See https://github.com/dantaki/SV2/wiki/input#snv-vcf for details'
sys.stderr.write(
'FATAL ERROR: SNV file(s) were not formatted correctly. See https://github.com/dantaki/SV2/wiki/input#snv-vcf for details\n'
)
sys.exit(1)
for snv in Snv:
if snv.id.get(bam.id) != None:
bam.Snv, bam.snv_index = snv, snv.id[bam.id]
if bam.id != None and bam.Snv == None:
sys.stderr.write(
'WARNING: BAM file {} sample name (@RG SM:<sample_id>):{} not found in SNV VCFs. Skipping {} ...\n'
.format(f, bam.id, f))
if bam.id == None:
sys.stderr.write(
'WARNING: Skipping BAM file {}. No sample name (@RG SM:<sample_id>). See https://github.com/dantaki/SV2/wiki/input#bam for details'
)
if bam.id != None and bam.Snv != None: bams.append(bam)
if len(bams) < 1:
print 'FATAL ERROR: BAM file(s) were not formatted correctly. See https://github.com/dantaki/SV2/wiki/input#bam for details'
sys.stderr.write(
'FATAL ERROR: BAM file(s) were not formatted correctly. See https://github.com/dantaki/SV2/wiki/input#bam for details\n'
)
sys.exit(1)
return bams
def write_config(self, hg19=None, hg38=None, mm10=None):
conf = ConfigParser()
for x in self.clfs['default']:
clf_fh = str(self.clfs['default'][x])
if not os.path.isfile(clf_fh):
realpath = get_path() + '/resources/training_sets/' + clf_fh
if os.path.isfile(realpath): self.clfs['default'][x] = realpath
else:
print 'FATAL ERROR: {} pickle file not found. If this file is missing, reinstall sv2: pip uninstall sv2 -y && pip install sv2-VERSION.tar.gz'.format(
realpath)
sys.stderr.write(
'FATAL ERROR: {} pickle file not found. If this file is missing, reinstall sv2: pip uninstall sv2 -y && pip install sv2-VERSION.tar.gz\n'
.format(realpath))
sys.exit(1)
dump_json(self.json, self.clfs)
if not os.path.isfile(self.fh):
conf_fh = open(self.fh, 'w')
conf.add_section('FASTA_PATHS')
conf.set('FASTA_PATHS', 'hg19', hg19)
conf.set('FASTA_PATHS', 'hg38', hg38)
conf.set('FASTA_PATHS', 'mm10', mm10)
conf.set('RESOURCE_DIR', 'sv2_resource', self.resource)
conf.write(conf_fh)
conf_fh.close()
else:
conf.read(self.fh)
if hg19 != None:
errFH(hg19)
conf.set('FASTA_PATHS', 'hg19', hg19)
if hg38 != None:
errFH(hg38)
conf.set('FASTA_PATHS', 'hg38', hg38)
if mm10 != None:
errFH(mm10)
conf.set('FASTA_PATHS', 'mm10', mm10)
if self.resource != None:
conf.set('RESOURCE_DIR', 'sv2_resource', self.resource)
with open(self.fh, 'w') as conf_fh:
conf.write(conf_fh)