def cli(ctx, bam, reference, genes_file, 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))
# 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"])
# Create an AACensus object
aa_census = AACensus(reference, mapped_read_collection_arr, genes, frames)
if output:
output.write(aa_census.coverage(frames))
output.close()
else:
click.echo(aa_census.coverage(frames))
def cli(ctx, bam, reference, bed4_file, output):
"""This script builds an amino acid census and returns its coverage.
The BAM alignment file corresponds to a pileup of sequences aligned to
the REFERENCE. A BAM index file (.bai) must also be present and, except
for the extension, have the same name as the BAM file. The REFERENCE must
be in FASTA format. The BED4_FILE must be a BED file with at least 4
columns and specify the gene locations within the REFERENCE.
The output is in CSV format."""
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))
# 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)
if output:
output.write(aa_census.coverage(frames))
output.close()
else:
click.echo(aa_census.coverage(frames))
class TestMappedRead:
@classmethod
def setup_class(self):
reference = TEST_PATH + "/data/hxb2_pol.fas"
bam = TEST_PATH + "/data/align.bam"
BED4_file = TEST_PATH + "/data/hxb2_pol.bed"
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))
genes = parse_BED4_file(BED4_file, rs[0].name)
# Determine which frames our genes are in
self.frames = set()
for gene in genes:
self.frames.add(genes[gene]["frame"])
self.aa_census = AACensus(reference, mapped_read_collection_arr, genes,
self.frames)
def test_coverage(self):
with open(VALID_COVERAGE_CSV, "r") as input:
coverage = input.read()
assert self.aa_census.coverage(self.frames) == coverage
def setup_class(self):
reference = TEST_PATH + "/data/hxb2_pol.fas"
bam = TEST_PATH + "/data/align.bam"
BED4_file = TEST_PATH + "/data/hxb2_pol.bed"
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))
genes = parse_BED4_file(BED4_file, rs[0].name)
# Determine which frames our genes are in
self.frames = set()
for gene in genes:
self.frames.add(genes[gene]["frame"])
self.aa_census = AACensus(reference, mapped_read_collection_arr, genes,
self.frames)
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 cli(ctx, bam, reference, variants, bed4_file, min_freq, mutation_db,
reporting_threshold, 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))
variants_obj = parse_nt_variants_from_vcf(variants, rs)
# 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('mf' + str(min_freq), 'freq<' + str(min_freq), True)
# Build the mutation database
mutation_db = MutationDB(mutation_db, genes)
# Generate the mutation report
if output:
output.write(
aa_vars.report_dr_mutations(mutation_db, reporting_threshold))
output.close()
else:
click.echo(
aa_vars.report_dr_mutations(mutation_db, reporting_threshold))
def aavar(bam, reference, variants, genes_file, min_freq, mutation_db, 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))
variants_obj = parse_nt_variants_from_vcf(variants, rs)
# 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_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'])
# Create an AACensus object
aa_census = AACensus(reference, mapped_read_collection_arr, genes, frames)
# Create AAVar collection and print the hmcf 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)
if output:
output.write(aa_vars.to_hmcf_file(CONFIDENT))
else:
click.echo(aa_vars.to_hmcf_file(CONFIDENT))
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):
reference = TEST_PATH + "/data/hxb2_pol.fas"
bam = TEST_PATH + "/data/align.bam"
genes_file = TEST_PATH + "/data/hxb2_pol.bed"
mutation_db = TEST_PATH + "/data/mutation_db.tsv"
min_freq = 0.01
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 = parse_nt_variants_from_vcf(VARIANTS_FILE, rs)
# 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_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'])
# Create an AACensus object
aa_census = AACensus(reference, mapped_read_collection_arr, genes,
frames)
# Find the AA mutations
self.aa_collection = AAVariantCollection.from_aacensus(aa_census)
# Build the mutation database
self.mutation_db = MutationDB(mutation_db, genes)
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)