def setUpClass(cls):
cls.dir = tempfile.mkdtemp()
# create a fasta file
cls.fa = os.path.join(cls.dir, 'genome.fa')
with open(cls.fa, mode='wt') as handle:
handle.write('>chr1\n')
handle.write('ACTGATGCTAGCTAGTATCTGACTCAGTAGCTCGAT\n')
# index the fasta file
fai = Faidx(cls.fa)
fai.close()
# set the final args that depend on the temp directory
get_options.set_attr('tempdir', cls.dir)
get_options.set_attr('reference', cls.fa)
outvcf = os.path.join(cls.dir, 'out.vcf.gz')
invcf = os.path.join(cls.dir, 'in.vcf.gz')
get_options.set_attr('vcf', invcf)
get_options.set_attr('out', outvcf)
# write a VCF to be converted. This includes one variant which cannot be
# converted. TODO: make a unit test to check for expected log output for
# unconvertible variant
with gzip.open(invcf, 'wt') as handle:
handle.write('##fileformat=VCFv4.1\n' \
'#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\n' \
'1\t10\t.\tT\tG\t100\tPASS\tAC=100\n' \
'1\t1000000\t.\tT\tG\t100\tPASS\tAC=100\n' \
'1\t2000000\t.\tA\tG\t100\tPASS\tAC=100\n')
def test_reindex_on_modification(self):
""" This test ensures that the index is regenerated when the FASTA
modification time is newer than the index modification time.
mdshw5/pyfaidx#50 """
faidx = Faidx('data/genes.fasta')
index_mtime = getmtime(faidx.indexname)
faidx.close()
os.utime('data/genes.fasta', (index_mtime + 10, ) * 2)
time.sleep(2)
faidx = Faidx('data/genes.fasta')
assert getmtime(faidx.indexname) > index_mtime
def test_reindex_on_modification(self):
""" This test ensures that the index is regenerated when the FASTA
modification time is newer than the index modification time.
mdshw5/pyfaidx#50 """
faidx = Faidx('data/genes.fasta')
index_mtime = getmtime(faidx.indexname)
faidx.close()
os.utime('data/genes.fasta', (index_mtime + 10, ) * 2)
time.sleep(2)
faidx = Faidx('data/genes.fasta')
assert getmtime(faidx.indexname) > index_mtime
class Genome(object):
def __init__(self, db):
from pyfaidx import Faidx
fa = os.path.join(app.config["DATA_FOLDER"], db, db + ".fa")
self.fasta = Faidx(fa)
def get_sequence(self, chr, start, end):
return self.fasta.fetch(chr, start, end)
def destroy(self):
self.fasta.close()
def setUpClass(cls):
cls.dir = tempfile.mkdtemp()
# create a fasta file
cls.fa = os.path.join(cls.dir, 'genome.fa')
with open(cls.fa, mode='wt') as handle:
handle.write('>chrN\n')
handle.write('NNTGATGCTAGCTAGTATCTG\n')
# index the fasta file
fai = Faidx(cls.fa)
fai.close()
def profile(self,
input_vcf_file,
ref_genome_file,
output_file,
raw_gt_format="GTR",
sample_id=None,
):
# unzip decompose and clean vcf file
clean_vcf_file = join_path(mkdtemp(), "clean.vcf")
if input_vcf_file.endswith(".gz"):
cmd = "gunzip -c " + input_vcf_file
else:
cmd = "cat " + input_vcf_file
cmd += " | vt decompose -s -"
cmd += " | grep -Pv \"\t\*\t\""
cmd += " | grep -v \"\\x3b\""
cmd += " | grep -v \"^M\""
cmd += " > " + clean_vcf_file
p, stdout_data = exec_sh(cmd, silent=True)
# parse the clean vcf file for the required fields
vcf_query_format = "'"
vcf_query_format += "%CHROM"
vcf_query_format += "\t%POS"
vcf_query_format += "\t%REF"
vcf_query_format += "\t%ALT"
vcf_query_format += "[\t%SAMPLE=%" + raw_gt_format + "]"
vcf_query_format += "\n"
vcf_query_format += "'"
cmd = "vcf-query"
cmd += " -f " + vcf_query_format
cmd += " " + clean_vcf_file
p, stdout_data = exec_sh(cmd, silent=True)
# get list of smaple id and prepare data structure
first_variant_record = stdout_data.decode('utf-8').split("\n")[0]
variant_items = first_variant_record.strip().split("\t")
samples_features = {}
for sample_idx in range(4, len(variant_items)):
gt_data = variant_items[sample_idx]
m = re.match(r"(?P<sample_id>.*)=(?P<raw_gt>.*)", gt_data)
sample_id = m.group("sample_id")
samples_features[sample_id] = copy.deepcopy(SNV_FEATURES_TEMPLATE)
# iterate over all vcf record and count variants for each sample
fa = Faidx(ref_genome_file)
for variant_record in stdout_data.decode('utf-8').split("\n"):
variant_items = variant_record.strip().split("\t")
if len(variant_items) < 4:
continue
chrom = variant_items[0]
pos = variant_items[1]
ref = variant_items[2]
alt = variant_items[3]
if len(ref) > 1:
continue
if ref == "-":
continue
if len(alt) > 1:
continue
if alt == "-":
continue
triplet = fa.fetch(chrom, int(pos)-1, int(pos)+1).seq
feature_id = SNV_FEATURES_HASH[ref][alt][triplet]
# iterate over all samples in the record
for sample_idx in range(4, len(variant_items)):
gt_data = variant_items[sample_idx]
m = re.match(r"(?P<sample_id>.*)=(?P<raw_gt>.*)", gt_data)
raw_gt = m.group("raw_gt")
if raw_gt == "0/0":
continue
samples_features[m.group("sample_id")][feature_id][FEATURE_QUANTITY] += 1
fa.close()
# write output feature file
with open(output_file, "w") as f_o:
header = VARIANT_TYPE
header += "\t" + VARIANT_SUBGROUP
header += "\t" + FEATURE_ID
for sample_id in samples_features:
header += "\t" + sample_id
f_o.write(header+"\n")
for feature_id in SNV_FEATURES_TEMPLATE:
feature_info = "{:s}\t{:s}\t{:s}".format(SNV_FEATURES_TEMPLATE[feature_id][VARIANT_TYPE],
SNV_FEATURES_TEMPLATE[feature_id][VARIANT_SUBGROUP],
feature_id,
)
for sample_id in samples_features:
feature_info += "\t" + str(samples_features[sample_id][feature_id][FEATURE_QUANTITY])
f_o.write(feature_info + "\n")
self.info()
self.info("Done!! The output file is at " + output_file)
