def setup_class(self):
self.references = parse_references_from_fasta('tests/data/ref1.fasta')
self.variant_collection = NTVariantCollection(self.references)
self.variant_collection.variants['ref1']['3']['t'] = NTVariant(
chrom='ref1',
pos=3,
ref='c',
alt='t',
qual=30,
info={
'DP': 400,
'AC': 12,
'AF': 0.03
})
self.variant_collection.variants['ref1']['10']['a'] = NTVariant(
chrom='ref1',
pos=10,
ref='a',
alt='t',
qual=23,
info={
'DP': 200,
'AC': 7,
'AF': 0.035
})
def ntvar(bam, reference, error_rate, output):
rs = parse_references_from_fasta(reference)
mapped_read_collection_arr = []
for r in rs:
# create MappedReadCollection object
mapped_read_collection_arr.append(parse_mapped_reads_from_bam(r, bam))
mapped_read_collection_arr = []
for r in rs:
# create MappedReadCollection object
mapped_read_collection_arr.append(parse_mapped_reads_from_bam(r, bam))
variants = NTVariantCollection.from_mapped_read_collections(
error_rate, rs, *mapped_read_collection_arr)
variants.filter('q30', 'QUAL<30', True)
variants.filter('ac5', 'AC<5', True)
variants.filter('dp100', 'DP<100', True)
if output:
output.write(variants.to_vcf_file())
output.close()
else:
click.echo(variants.to_vcf_file())
def parse_nt_variants_from_vcf(variant_file, references):
"""Build variants object from a vcf file"""
obj = NTVariantCollection(references)
# Read in and parse the variants file
# The file uses 1 as the first position but 0 is the first position in
# mapped reads.
with open(variant_file, "r") as input:
for line in input:
if line[0] != "#":
chrom, pos, var_id, ref, alt, qual, var_filter, info = \
line.rstrip().split("\t")
dp, ac, af = info.rstrip().split(';')
dp = dp.rstrip().split('=')[1]
ac = ac.rstrip().split('=')[1]
af = af.rstrip().split('=')[1]
variant_obj = NTVariant(chrom=chrom,
id=var_id,
pos=int(pos),
ref=ref,
alt=alt,
qual=qual,
filter=var_filter,
info={
'DP': dp,
'AC': ac,
'AF': af
})
obj.variants[chrom][int(pos)][alt] = variant_obj
return obj
def test_from_aacensus(self):
bam = TEST_PATH + "/data/align.bam"
BED4_file = TEST_PATH + "/data/hxb2_pol.bed"
mapped_read_collection_arr = []
error_rate = 0.0038
# Create a MappedReadCollection object
for r in self.references:
mapped_read_collection_arr.append(
parse_mapped_reads_from_bam(r, bam))
variants = NTVariantCollection.from_mapped_read_collections(
error_rate, self.references, *mapped_read_collection_arr)
variants.filter('q30', 'QUAL<30', True)
variants.filter('ac5', 'AC<5', True)
variants.filter('dp100', 'DP<100', True)
# Mask the unconfident differences
for mrc in mapped_read_collection_arr:
mrc.mask_unconfident_differences(variants)
# Parse the genes from the gene file
genes = parse_BED4_file(BED4_file, self.references[0].name)
# Determine which frames our genes are in
frames = set()
for gene in genes:
frames.add(genes[gene]['frame'])
aa_census = AACensus(self.reference, mapped_read_collection_arr, genes,
frames)
test_variants = CodonVariantCollection.from_aacensus(aa_census)
ref_seq = self.references[0].seq
for gene in test_variants.variants:
assert gene in genes
for pos in test_variants.variants[gene]:
for frame in frames:
nt_pos = pos / 3 - frame
assert nt_pos >= genes[gene]['start'] or nt_pos <= genes[
gene]['end']
for codon in test_variants.variants[gene][pos]:
ref_codon = ref_seq[(pos):(pos) + 3].lower()
assert codon != ref_codon
def codonvar(bam, reference, offset, bed4_file, variants, error_rate, output):
rs = parse_references_from_fasta(reference)
mapped_read_collection_arr = []
# Create a MappedReadCollection object
for r in rs:
mapped_read_collection_arr.append(parse_mapped_reads_from_bam(r, bam))
if variants:
variants_obj = parse_nt_variants_from_vcf(variants, rs)
else:
variants = NTVariantCollection.from_mapped_read_collections(
error_rate, rs, *mapped_read_collection_arr)
variants.filter('q30', 'QUAL<30', True)
variants.filter('ac5', 'AC<5', True)
variants.filter('dp100', 'DP<100', True)
variants_obj = variants
# Mask the unconfident differences
for mrc in mapped_read_collection_arr:
mrc.mask_unconfident_differences(variants_obj)
# Parse the genes from the gene file
genes = parse_BED4_file(bed4_file, rs[0].name)
# Determine which frames our genes are in
frames = set()
for gene in genes:
frames.add(genes[gene]['frame'])
aa_census = AACensus(reference, mapped_read_collection_arr, genes, frames)
codon_variants = CodonVariantCollection.from_aacensus(aa_census)
if output:
output.write(codon_variants.to_csv_file(offset))
output.close()
else:
click.echo(codon_variants.to_csv_file(offset))
def setup(self):
bam = TEST_PATH + "/data/align.bam"
reference = TEST_PATH + "/data/hxb2_pol.fas"
genes_file = TEST_PATH + "/data/hxb2_pol.bed"
error_rate = 0.0038
rs = parse_references_from_fasta(reference)
mapped_read_collection_arr = []
# Create a MappedReadCollection object
for r in rs:
mapped_read_collection_arr.append(
parse_mapped_reads_from_bam(r, bam))
variants = NTVariantCollection.from_mapped_read_collections(
error_rate, rs, *mapped_read_collection_arr)
variants.filter('q30', 'QUAL<30', True)
variants.filter('ac5', 'AC<5', True)
variants.filter('dp100', 'DP<100', True)
# Mask the unconfident differences
for mrc in mapped_read_collection_arr:
mrc.mask_unconfident_differences(variants)
# Parse the genes from the gene file
genes = parse_genes_file(genes_file, rs[0].name)
# Determine which frames our genes are in
frames = set()
for gene in genes:
frames.add(genes[gene]['frame'])
aa_census = AACensus(reference, mapped_read_collection_arr, genes,
frames)
self.codon_variants = CodonVariantCollection.from_aacensus(aa_census)
def aavar(bam, reference, bed4_file, variants, mutation_db, min_freq,
error_rate, output):
rs = parse_references_from_fasta(reference)
mapped_read_collection_arr = []
for r in rs:
# Create a MappedReadCollection object
mapped_read_collection_arr.append(parse_mapped_reads_from_bam(r, bam))
if variants:
variants_obj = parse_nt_variants_from_vcf(variants, rs)
else:
variants = NTVariantCollection.from_mapped_read_collections(
error_rate, rs, *mapped_read_collection_arr)
variants.filter('q30', 'QUAL<30', True)
variants.filter('ac5', 'AC<5', True)
variants.filter('dp100', 'DP<100', True)
variants_obj = variants
# Mask the unconfident differences
for mrc in mapped_read_collection_arr:
mrc.mask_unconfident_differences(variants_obj)
# Parse the genes from the gene file
genes = parse_BED4_file(bed4_file, rs[0].name)
# Determine which frames our genes are in
frames = set()
for gene in genes:
frames.add(genes[gene]['frame'])
# Create an AACensus object
aa_census = AACensus(reference, mapped_read_collection_arr, genes, frames)
# Create AAVar collection and print the aavf file
aa_vars = AAVariantCollection.from_aacensus(aa_census)
# Filter for mutant frequency
aa_vars.filter('mf0.01', 'freq<0.01', True)
# Build the mutation database and update collection
if mutation_db is not None:
mutation_db = MutationDB(mutation_db, genes)
aa_vars.apply_mutation_db(mutation_db)
aavf_obj = aa_vars.to_aavf_obj("aavar", os.path.basename(reference),
CONFIDENT)
records = list(aavf_obj)
if output:
writer = parser.Writer(output, aavf_obj)
else:
writer = parser.Writer(sys.stdout, aavf_obj)
for record in records:
writer.write_record(record)
if output:
output.close
writer.close()
def analyze_reads(self, fasta_id, variant_filters, reporting_threshold,
generate_consensus):
# Map reads against reference using bowtietwo
if not self.quiet:
print("# Mapping reads...")
try:
bam = self.generate_bam(fasta_id)
except Exception as error:
raise (error)
if not self.quiet:
print("# Loading read mappings...")
# cmd_consensus
if generate_consensus:
cons_seq_file = open("%s/consensus.fasta" % self.output_dir, "w+")
mapped_read_collection_arr = []
for r in self.references:
mrc = parse_mapped_reads_from_bam(r, bam)
mapped_read_collection_arr.append(mrc)
consensus_seq = mrc.to_consensus(self.consensus_pct)
if generate_consensus and len(consensus_seq) > 0:
cons_seq_file.write('>{0}_{1}_{2}\n{3}'.format(
fasta_id, reporting_threshold, r.name, consensus_seq))
if generate_consensus:
cons_seq_file.close()
# cmd_callntvar
if not self.quiet:
print("# Identifying variants...")
variants = NTVariantCollection.from_mapped_read_collections(
variant_filters[ERROR_RATE], self.references,
*mapped_read_collection_arr)
variants.filter('q%s' % variant_filters[MIN_VARIANT_QUAL],
'QUAL<%s' % variant_filters[MIN_VARIANT_QUAL], True)
variants.filter('ac%s' % variant_filters[MIN_AC],
'AC<%s' % variant_filters[MIN_AC], True)
variants.filter('dp%s' % variant_filters[MIN_DP],
'DP<%s' % variant_filters[MIN_DP], True)
vcf_file = open("%s/hydra.vcf" % self.output_dir, "w+")
vcf_file.write(variants.to_vcf_file())
vcf_file.close()
# cmd_aa_census
if not self.quiet:
print("# Masking filtered variants...")
for mrc in mapped_read_collection_arr:
mrc.mask_unconfident_differences(variants)
if not self.quiet:
print("# Building amino acid census...")
# Determine which frames our genes are in
frames = set()
for gene in self.genes:
frames.add(self.genes[gene]['frame'])
aa_census = AACensus(self.reference, mapped_read_collection_arr,
self.genes, frames)
coverage_file = open("%s/coverage_file.csv" % self.output_dir, "w+")
coverage_file.write(aa_census.coverage(frames))
coverage_file.close()
# cmd_aavariants
if not self.quiet:
print("# Finding amino acid mutations...")
# Create AAVar collection and print the aavf file
aa_vars = AAVariantCollection.from_aacensus(aa_census)
# Filter for mutant frequency
aa_vars.filter('mf%s' % variant_filters[MIN_FREQ],
'freq<%s' % variant_filters[MIN_FREQ], True)
# Build the mutation database and update collection
if self.mutation_db is not None:
mutation_db = MutationDB(self.mutation_db, self.genes)
aa_vars.apply_mutation_db(mutation_db)
aavf_obj = aa_vars.to_aavf_obj("hydra",
os.path.basename(self.reference),
CONFIDENT)
records = list(aavf_obj)
mut_report = open("%s/mutation_report.aavf" % self.output_dir, "w+")
writer = parser.Writer(mut_report, aavf_obj)
for record in records:
writer.write_record(record)
mut_report.close()
writer.close()
# cmd_drmutations
if not self.quiet:
print("# Writing drug resistant mutation report...")
dr_report = open("%s/dr_report.csv" % self.output_dir, "w+")
dr_report.write(
aa_vars.report_dr_mutations(mutation_db, reporting_threshold))
dr_report.close()
self.output_stats(mapped_read_collection_arr)
def analyze_reads(self, fasta_id, filters, reporting_threshold,
generate_consensus):
# Map reads against reference using bowtietwo
if not self.quiet:
print("# Mapping reads...")
bam = self.generate_bam(fasta_id)
if not self.quiet:
print("# Loading read mappings...")
# cmd_consensus
if generate_consensus:
cons_seq_file = open("%s/consensus.fasta" % self.output_dir, "w+")
mapped_read_collection_arr = []
for r in self.references:
mrc = parse_mapped_reads_from_bam(r, bam)
mapped_read_collection_arr.append(mrc)
if generate_consensus:
cons_seq_file.write('>{0}_{1}_{2}\n{3}'.format(
fasta_id, reporting_threshold, r.name,
mrc.to_consensus(self.consensus_pct)))
if generate_consensus:
cons_seq_file.close()
# cmd_callntvar
if not self.quiet:
print("# Identifying variants...")
variants = NTVariantCollection.from_mapped_read_collections(
filters["error_rate"], self.references,
*mapped_read_collection_arr)
variants.filter('q%s' % filters["min_qual"],
'QUAL<%s' % filters["min_qual"], True)
variants.filter('ac%s' % filters["min_ac"],
'AC<%s' % filters["min_ac"], True)
variants.filter('dp%s' % filters["min_dp"],
'DP<%s' % filters["min_dp"], True)
vcf_file = open("%s/hydra.vcf" % self.output_dir, "w+")
vcf_file.write(variants.to_vcf_file())
vcf_file.close()
# cmd_aa_census
if not self.quiet:
print("# Masking filtered variants...")
for mrc in mapped_read_collection_arr:
mrc.mask_unconfident_differences(variants)
if not self.quiet:
print("# Building amino acid census...")
# Determine which frames our genes are in
frames = set()
for gene in self.genes:
frames.add(self.genes[gene]['frame'])
aa_census = AACensus(self.reference, mapped_read_collection_arr,
self.genes, frames)
coverage_file = open("%s/coverage_file.csv" % self.output_dir, "w+")
coverage_file.write(aa_census.coverage(frames))
coverage_file.close()
# cmd_aavariants
if not self.quiet:
print("# Finding amino acid mutations...")
# Create AAVar collection and print the hmcf file
aa_vars = AAVariantCollection.from_aacensus(aa_census)
# Filter for mutant frequency
aa_vars.filter('mf%s' % filters['min_freq'],
'freq<%s' % filters['min_freq'], True)
# Build the mutation database and update collection
if self.mutation_db is not None:
mutation_db = MutationDB(self.mutation_db, self.genes)
aa_vars.apply_mutation_db(mutation_db)
mut_report = open("%s/mutation_report.hmcf" % self.output_dir, "w+")
mut_report.write(aa_vars.to_hmcf_file(CONFIDENT))
mut_report.close()
# cmd_drmutations
if not self.quiet:
print("# Writing drug resistant mutation report...")
dr_report = open("%s/dr_report.csv" % self.output_dir, "w+")
dr_report.write(aa_vars.report_dr_mutations(mutation_db,
reporting_threshold))
dr_report.close()
self.output_stats(mapped_read_collection_arr)
class TestNTVariantCollection:
@classmethod
def setup_class(self):
self.references = parse_references_from_fasta('tests/data/ref1.fasta')
self.variant_collection = NTVariantCollection(self.references)
self.variant_collection.variants['ref1']['3']['t'] = NTVariant(
chrom='ref1',
pos=3,
ref='c',
alt='t',
qual=30,
info={
'DP': 400,
'AC': 12,
'AF': 0.03
})
self.variant_collection.variants['ref1']['10']['a'] = NTVariant(
chrom='ref1',
pos=10,
ref='a',
alt='t',
qual=23,
info={
'DP': 200,
'AC': 7,
'AF': 0.035
})
def test_from_mapped_read_collections(self):
#TODO: add actual test for Variants.from_mapped_reads method
assert True
def test_to_vcf_file(self):
vcf_file_string = self.variant_collection.to_vcf_file()
assert "##fileformat=VCFv4.2" in vcf_file_string
assert "##fileDate=" in vcf_file_string
assert "##source=quasitools" in vcf_file_string
assert "##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">" in vcf_file_string
assert "##INFO=<ID=AC,Number=A,Type=Integer,Description=\"Allele Count\">" in vcf_file_string
assert "##INFO=<ID=AF,Number=A,Type=Float,Description=\"Allele Frequency\">" in vcf_file_string
assert "#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO" in vcf_file_string
assert "ref1\t3\t.\tc\tt\t30\t.\tDP=400;AC=12;AF=0.0300" in vcf_file_string
assert "ref1\t10\t.\ta\tt\t23\t.\tDP=200;AC=7;AF=0.0350" in vcf_file_string
def test_calculate_variant_qual(self):
qual1 = self.variant_collection._NTVariantCollection__calculate_variant_qual(
0.01, 12, 400)
qual2 = self.variant_collection._NTVariantCollection__calculate_variant_qual(
0.01, 7, 200)
assert qual1 == 30
assert qual2 == 23
def test_filter(self):
variant_collection_copy = copy.deepcopy(self.variant_collection)
variant_collection_copy.filter('q30', 'QUAL<30', True)
assert variant_collection_copy.variants['ref1']['3'][
't'].filter == 'PASS'
assert variant_collection_copy.variants['ref1']['10'][
'a'].filter == 'q30'
variant_collection_copy.filter('ac5', 'AC<5', True)
assert variant_collection_copy.variants['ref1']['3'][
't'].filter == 'PASS'
assert variant_collection_copy.variants['ref1']['10'][
'a'].filter == 'q30'
variant_collection_copy.filter('dp100', 'DP<100', True)
assert variant_collection_copy.variants['ref1']['3'][
't'].filter == 'PASS'
assert variant_collection_copy.variants['ref1']['10'][
'a'].filter == 'q30'
def test_valid_vcf_file(self):
"""Tests to ensure that valid vcf files are parsed properly."""
reference = TEST_PATH + \
"/data/hxb2_pol.fas"
bam = TEST_PATH + "/data/align.bam"
rs = parse_references_from_fasta(reference)
mapped_read_collection_arr = []
for r in rs:
# Create a MappedReadCollection object
mapped_read_collection_arr.append(
parse_mapped_reads_from_bam(r, bam))
variants_obj = NTVariantCollection(rs)
for i in range(0, 20):
variant = NTVariant(chrom="hxb2_pol",
pos=i,
id=".",
ref='a',
alt='t',
qual="50",
filter="PASS",
info={
"DP": "300",
"AC": "1",
"AF": "0.0025"
})
variants_obj.variants["hxb2_pol"][i]['t'] = variant
#Create a valid vcf file
valid_vcf_file = TEST_PATH + "/data/valid_vcf_file.vcf"
with open(valid_vcf_file, "w+") as f:
f.write(
"##fileformat=VCFv4.2\n"
"##fileDate=20171005\n"
"##source=quasitools\n"
"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">\n"
"##INFO=<ID=AC,Number=A,Type=Integer,Description=\"Allele Count\">\n"
"##INFO=<ID=AF,Number=A,Type=Float,Description=\"Allele Frequency\">\n"
"##FILTER=<ID=q30,Description=\"Quality below 30\">\n"
"##FILTER=<ID=dp100,Description=\"Read depth below 100\">\n"
"##FILTER=<ID=ac5,Description=\"Allele count below 5\">\n"
"#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO")
for rid in variants_obj.variants:
for pos in variants_obj.variants[rid]:
for alt in variants_obj.variants[rid][pos]:
variant = variants_obj.variants[rid][pos][alt]
f.write("\n%s\t%i\t%s\t%s\t%s\t%s\t%s" %
(variant.chrom, int(
variant.pos), variant.id, variant.ref,
variant.alt, variant.qual, variant.filter))
f.write(
"\tDP=%i;AC=%i;AF=%0.4f" %
(int(variant.info["DP"]), int(variant.info["AC"]),
float(variant.info["AF"])))
parsed_nt_var = parse_nt_variants_from_vcf(valid_vcf_file, rs)
# Check equality of parsed NTVariantCollection vs. the valid NTVariantCollection
for rid in parsed_nt_var.variants:
for pos in parsed_nt_var.variants[rid]:
for alt in parsed_nt_var.variants[rid][pos]:
parsed_variant = parsed_nt_var.variants[rid][pos][alt]
variant = variants_obj.variants[rid][pos][alt]
assert parsed_variant.chrom == variant.chrom
assert parsed_variant.pos == variant.pos
assert parsed_variant.id == variant.id
assert parsed_variant.ref == variant.ref
assert parsed_variant.alt == variant.alt
assert parsed_variant.qual == variant.qual
assert parsed_variant.filter == variant.filter
assert parsed_variant.info["DP"] == variant.info["DP"]
assert parsed_variant.info["AC"] == variant.info["AC"]
assert parsed_variant.info["AF"] == variant.info["AF"]
os.remove(valid_vcf_file)