def call_sites(args):
"""Find the sites with SNPs in a sample.
The sample alignment is sorted, duplicate reads are removed, a pileup is generated, and
snps are called.
This function expects, or creates '(*)', the following files arranged
in the following way:
reference
referenceFile.fasta
samples
sample_name_one/reads.sam
sample_name_one/reads.unsorted.bam*
sample_name_one/reads.sorted.bam*
sample_name_one/reads.sorted.deduped.bam*
sample_name_one/reads.all.pileup*
sample_name_one/var.flt.vcf*
The input files are created outside of this function. The package
documentation provides an example of preparing these files based on the
lambda_virus sequence that is used as one test for this package.
Parameters
----------
args : argparse.Namespace
referenceFile : File path of the reference fasta file
sampleDir : Relative or absolute directory of the sample
"""
utils.print_log_header(classpath=True)
utils.print_arguments(args)
#==========================================================================
# Validate inputs
#==========================================================================
# Verify reference fasta file exists and is not empty
reference_file_path = args.referenceFile
utils.verify_non_empty_input_files("Reference file", [reference_file_path],
error_handler="global")
sample_dir = args.sampleDir
sam_file = os.path.join(sample_dir, "reads.sam")
utils.verify_non_empty_input_files("Sample SAM file", [sam_file],
error_handler="sample")
sample_id = utils.sample_id_from_dir(sample_dir)
#==========================================================================
# Convert sam to bam file, selecting only the mapped reads
#==========================================================================
# Check for fresh bam file; if not, convert to bam file with only mapped reads
unsorted_bam_file = os.path.join(sample_dir, "reads.unsorted.bam")
needs_rebuild = utils.target_needs_rebuild([sam_file], unsorted_bam_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# Unsorted bam file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
version_str = utils.extract_version_str("SAMtools",
"samtools 2>&1 > /dev/null")
# Substitute the default parameters if the user did not specify samtools view parameters
samtools_samfilter_params = os.environ.get(
"SamtoolsSamFilter_ExtraParams") or "-F 4"
command_line = "samtools view -S -b " + samtools_samfilter_params + " -o " + unsorted_bam_file + ' ' + sam_file
verbose_print(
"# Convert sam file to bam file with only mapped positions.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, sys.stdout)
utils.sample_error_on_missing_file(unsorted_bam_file, "samtools view")
verbose_print("")
#==========================================================================
# Sort the BAM file
#==========================================================================
# Check for fresh sorted bam file; if not, sort it
sorted_bam_file = os.path.join(sample_dir, "reads.sorted.bam")
needs_rebuild = utils.target_needs_rebuild([unsorted_bam_file],
sorted_bam_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# Sorted bam file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
version_str = utils.extract_version_str("SAMtools",
"samtools 2>&1 > /dev/null")
samtools_sort_extra_params = os.environ.get(
"SamtoolsSort_ExtraParams") or ""
# Inspect the samtools version to determine how to execute samtools
# Use the -o FILE command line option with SAMtools 1.3 and higher
samtools_version = version_str.split()[-1] # just the number
if samtools_version < "1.3":
command_line = "samtools sort " + samtools_sort_extra_params + ' ' + unsorted_bam_file + ' ' + os.path.join(
sample_dir, "reads.sorted")
else:
command_line = "samtools sort " + samtools_sort_extra_params + " -o " + sorted_bam_file + ' ' + unsorted_bam_file
verbose_print("# Convert bam to sorted bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, sys.stdout)
utils.sample_error_on_missing_file(sorted_bam_file, "samtools sort")
verbose_print("")
#==========================================================================
# Mark duplicate reads, so they will be ignored in subsequent steps
#==========================================================================
remove_duplicate_reads = os.environ.get("RemoveDuplicateReads") or "true"
remove_duplicate_reads = remove_duplicate_reads.lower()
if remove_duplicate_reads == "true":
# Check for fresh deduped bam file; if not, remove duplicate reads
deduped_bam_file = os.path.join(sample_dir, "reads.sorted.deduped.bam")
needs_rebuild = utils.target_needs_rebuild([sorted_bam_file],
deduped_bam_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# Deduped bam file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
classpath = os.environ.get("CLASSPATH")
if not classpath or "picard" not in classpath.lower():
utils.global_error(
"Error: cannot execute Picard. Define the path to Picard in the CLASSPATH environment variable."
)
else:
version_str = utils.extract_version_str(
"Picard",
"java picard.cmdline.PicardCommandLine MarkDuplicates --version 2>&1"
)
picard_jvm_extra_params = os.environ.get(
"PicardJvm_ExtraParams") or ""
picard_mark_duplicates_extra_params = os.environ.get(
"PicardMarkDuplicates_ExtraParams") or ""
tmpdir = os.environ.get("TMPDIR") or os.environ.get("TMP_DIR")
tmp_option = " TMP_DIR=" + tmpdir if tmpdir else ""
command_line = "java " + picard_jvm_extra_params + ' ' + "picard.cmdline.PicardCommandLine MarkDuplicates INPUT=" + sorted_bam_file + " OUTPUT=" + deduped_bam_file + " METRICS_FILE=" + os.path.join(
sample_dir, "duplicate_reads_metrics.txt"
) + tmp_option + ' ' + picard_mark_duplicates_extra_params
verbose_print("# Remove duplicate reads from bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, sys.stdout)
utils.sample_error_on_missing_file(deduped_bam_file,
"picard MarkDuplicates")
verbose_print("")
pileup_input_file = deduped_bam_file
else:
pileup_input_file = sorted_bam_file
#==========================================================================
# Create the pileup file
#==========================================================================
# Check for fresh pileup; if not, create it
pileup_file = os.path.join(sample_dir, "reads.all.pileup")
needs_rebuild = utils.target_needs_rebuild(
[pileup_input_file, reference_file_path], pileup_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# Pileup file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
version_str = utils.extract_version_str("SAMtools",
"samtools 2>&1 > /dev/null")
samtools_mpileup_extra_params = os.environ.get(
"SamtoolsMpileup_ExtraParams") or ""
command_line = "samtools mpileup " + samtools_mpileup_extra_params + " -f " + reference_file_path + ' ' + pileup_input_file
verbose_print("# Create pileup from bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, pileup_file)
utils.sample_error_on_missing_file(pileup_file, "samtools mpileup")
verbose_print("")
#==========================================================================
# Find the sites with SNPs
#==========================================================================
# Check for fresh unfiltered vcf; if not, create it
vcf_file = os.path.join(sample_dir, "var.flt.vcf")
needs_rebuild = utils.target_needs_rebuild([pileup_file], vcf_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# VCF file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
classpath = os.environ.get("CLASSPATH")
if not classpath or "varscan" not in classpath.lower():
utils.global_error(
"Error: cannot execute VarScan. Define the path to VarScan in the CLASSPATH environment variable."
)
else:
version_str = utils.extract_version_str(
"VarScan",
"java net.sf.varscan.VarScan 2>&1 > /dev/null | head -n 1 | cut -d ' ' -f 2"
)
varscan_jvm_extra_params = os.environ.get(
"VarscanJvm_ExtraParams") or ""
varscan_mpileup2snp_extra_params = os.environ.get(
"VarscanMpileup2snp_ExtraParams") or ""
command_line = "java " + varscan_jvm_extra_params + " net.sf.varscan.VarScan mpileup2snp " + pileup_file + " --output-vcf 1 " + varscan_mpileup2snp_extra_params
verbose_print("# Create vcf file")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, vcf_file)
utils.sample_error_on_missing_file(vcf_file, "VarScan")
utils.sample_error_on_file_contains(vcf_file, "OutOfMemoryError",
"VarScan")
utils.sample_error_on_file_contains(vcf_file, "Insufficient",
"VarScan")
def call_sites(args):
"""Find the sites with SNPs in a sample.
The sample alignment is sorted, duplicate reads are removed, a pileup is generated, and
snps are called.
This function expects, or creates '(*)', the following files arranged
in the following way:
reference
referenceFile.fasta
samples
sample_name_one/reads.sorted.deduped.indelrealigned.bam
sample_name_one/reads.all.pileup*
sample_name_one/var.flt.vcf*
The input files are created outside of this function. The package
documentation provides an example of preparing these files based on the
lambda_virus sequence that is used as one test for this package.
Parameters
----------
args : argparse.Namespace
referenceFile : File path of the reference fasta file
sampleDir : Relative or absolute directory of the sample
"""
utils.print_log_header(classpath=True)
utils.print_arguments(args)
#==========================================================================
# Validate inputs
#==========================================================================
# Verify reference fasta file exists and is not empty
reference_file_path = args.referenceFile
utils.verify_non_empty_input_files("Reference file", [reference_file_path],
error_handler="global")
sample_dir = args.sampleDir
remove_duplicate_reads = os.environ.get("RemoveDuplicateReads",
"true").lower() == "true"
enable_local_realignment = os.environ.get("EnableLocalRealignment",
"true").lower() == "true"
input_bam_file = os.path.join(sample_dir, "reads.sorted.bam")
input_bam_file = utils.add_file_suffix(input_bam_file,
".deduped",
enable=remove_duplicate_reads)
input_bam_file = utils.add_file_suffix(input_bam_file,
".indelrealigned",
enable=enable_local_realignment)
utils.verify_non_empty_input_files("Sample BAM file", [input_bam_file],
error_handler="sample")
sample_id = utils.sample_id_from_dir(sample_dir)
#==========================================================================
# Create the pileup file
#==========================================================================
# Check for fresh pileup; if not, create it
pileup_file = os.path.join(sample_dir, "reads.all.pileup")
needs_rebuild = utils.target_needs_rebuild(
[input_bam_file, reference_file_path], pileup_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# Pileup file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
version_str = utils.extract_version_str("SAMtools",
"samtools 2>&1 > /dev/null")
samtools_mpileup_extra_params = os.environ.get(
"SamtoolsMpileup_ExtraParams") or ""
command_line = "samtools mpileup " + samtools_mpileup_extra_params + " -f " + reference_file_path + ' ' + input_bam_file
verbose_print("# Create pileup from bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, pileup_file)
utils.sample_error_on_missing_file(pileup_file, "samtools mpileup")
verbose_print("")
#==========================================================================
# Find the sites with SNPs
#==========================================================================
# Check for fresh unfiltered vcf; if not, create it
vcf_file = os.path.join(sample_dir, "var.flt.vcf")
needs_rebuild = utils.target_needs_rebuild([pileup_file], vcf_file)
if not args.forceFlag and not needs_rebuild:
verbose_print(
"# VCF file is already freshly created for %s. Use the -f option to force a rebuild."
% sample_id)
else:
jar_file_path = utils.find_path_in_path_list("VarScan", "CLASSPATH")
if not jar_file_path:
utils.global_error(
"Error: cannot execute VarScan. Define the path to VarScan.jar in the CLASSPATH environment variable."
)
else:
version_str = utils.extract_version_str(
"VarScan", "java -jar " + jar_file_path +
" 2>&1 > /dev/null | head -n 1 | cut -d ' ' -f 2")
varscan_jvm_extra_params = os.environ.get(
"VarscanJvm_ExtraParams") or ""
varscan_mpileup2snp_extra_params = os.environ.get(
"VarscanMpileup2snp_ExtraParams") or ""
command_line = "java " + varscan_jvm_extra_params + " -jar " + jar_file_path + " mpileup2snp " + pileup_file + " --output-vcf 1 " + varscan_mpileup2snp_extra_params
verbose_print("# Create vcf file")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, vcf_file)
utils.sample_error_on_missing_file(vcf_file, "VarScan")
utils.sample_error_on_file_contains(vcf_file, "OutOfMemoryError",
"VarScan")
utils.sample_error_on_file_contains(vcf_file, "Insufficient",
"VarScan")
def collect_metrics(args):
"""Collect the quality metrics and SNP metrics for a sample.
This function expects, or creates '(*)', the following files arranged
in the following way:
reference
referenceFile.fasta
samples
sample_name_one/*.fastq.gz
sample_name_one/reads.sam
sample_name_one/reads.sorted.deduped.bam
sample_name_one/reads.sorted.bam
sample_name_one/reads.all.pileup
sample_name_one/var.flt.vcf
sample_name_one/var.flt_preserved.vcf
sample_name_one/consensus.fasta
sample_name_one/consensus_preserved.fasta
sample_name_one/consensus.vcf
sample_name_one/consensus_preserved.vcf
sample_name_one/metrics*
The input files are created outside of this function. The package
documentation provides an example of preparing these files based on the
lambda_virus sequence that is used as one test for this package.
Parameters
----------
args : argparse.Namespace
referenceFile : File path of the reference fasta file
sampleDir : Relative or absolute directory of the sample
consensusFastaFileName : File name of the consensus fasta file which must exist in the sample directory
consensusPreservedFastaFileName : File name of the consensus preserved fasta file which must exist in the sample directory
consensusVcfFileName : File name of the consensus vcf file which must exist in the sample directory
consensusPreservedVcfFileName : File name of the consensus preserved vcf file which must exist in the sample directory
maxSnps : Maximum allowed number of SNPs per sample
metricsFile : Output file. Relative or absolute path to the metrics file
"""
utils.print_log_header(classpath=True)
utils.print_arguments(args)
#==========================================================================
# Validate inputs
#==========================================================================
# Verify reference fasta file exists and is not empty
reference_file_path = args.referenceFile
utils.verify_non_empty_input_files("Reference file", [reference_file_path], error_handler="global")
sample_dir = args.sampleDir
utils.verify_non_empty_directory("Sample directory", sample_dir, error_handler="sample", continue_possible=False)
metrics_file_path = args.metricsFile
max_allowed_snps = args.maxSnps
consensus_vcf_file_name = args.consensusVcfFileName
consensus_preserved_vcf_file_name = args.consensusPreservedVcfFileName
consensus_fasta_file_name = args.consensusFastaFileName
consensus_preserved_fasta_file_name = args.consensusPreservedFastaFileName
sample_id = utils.sample_id_from_dir(sample_dir)
#==========================================================================
# Read existing metrics file so some metrics can be reused
#==========================================================================
try:
metrics = utils.read_properties(metrics_file_path)
except IOError:
metrics = dict()
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Get machine and flowcell from fastq header"))
#-------------------------
machine = ""
flowcell = ""
fastq_files = fastq.list_fastq_files(sample_dir)
fastq_files = [f for f in fastq_files if os.path.isfile(f)] # Exclude broken symlinks
if not fastq_files:
handle_error("No fastq files were found.")
else:
tags = fastq.extract_metadata_tags(fastq_files[0])
if tags:
machine = tags.instrument or ""
flowcell = tags.flow_cell or ""
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Sum file sizes of paired fastq files"))
#-------------------------
fastq_file_size = ""
fastq_file_list = ""
if fastq_files:
fastq_file_size = sum([os.path.getsize(file) for file in fastq_files])
# Make a comma separated list of just the fastq file names without directories
fastq_file_list = [os.path.basename(file) for file in fastq_files]
fastq_file_list = ", ".join(fastq_file_list)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Calculate number of reads and %mapped from sam file"))
#-------------------------
num_reads = ""
percent_reads_mapped = ""
file = os.path.join(sample_dir, "reads.sam")
if verify_input_file("SAM file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
num_reads = metrics.get("numberReads", "") # reuse already fresh metrics
percent_reads_mapped = metrics.get("percentReadsMapped", "") # reuse already fresh metrics
if num_reads and percent_reads_mapped:
verbose_print("Reusing previously calculated number of reads and %mapped")
else:
num_reads = command.run("samtools view -S -c " + file)
num_reads = num_reads.strip()
mapped = command.run("samtools view -S -c -F 4 " + file)
mapped = mapped.strip()
try:
percent_reads_mapped = 100.0 * float(mapped) / float(num_reads)
percent_reads_mapped = "%.2f" % percent_reads_mapped
except ValueError:
handle_error("Cannot calculate number of reads and %mapped.")
#-------------------------
# Calculate number of duplicate reads from deduped bam file
#-------------------------
num_dup_reads = ""
remove_duplicate_reads = os.environ.get("RemoveDuplicateReads") or "true"
remove_duplicate_reads = remove_duplicate_reads.lower()
if remove_duplicate_reads == "true":
verbose_print("# %s %s" % (utils.timestamp(), "Calculate number of duplicate reads from deduped bam file"))
file = os.path.join(sample_dir, "reads.sorted.deduped.bam")
if verify_input_file("Deduped BAM file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
num_dup_reads = metrics.get("numberDupReads", "") # reuse already fresh metrics
if num_dup_reads:
verbose_print("Reusing previously calculated number of duplicate reads")
else:
num_dup_reads = command.run("samtools view -S -c -f 1024 " + file)
num_dup_reads = num_dup_reads.strip()
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Calculate mean insert size from bam file"))
#-------------------------
ave_insert_size = ""
file = os.path.join(sample_dir, "reads.sorted.bam")
if verify_input_file("BAM file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
ave_insert_size = metrics.get("aveInsertSize", "") # reuse already fresh metrics
if ave_insert_size:
verbose_print("Reusing previously calculated mean insert size")
else:
# Extract inferred insert sizes (TLEN, column 9 of BAM file) for reads "mapped in proper pair" (2) and "first in pair" (64) = 66
tempfile = NamedTemporaryFile(delete=False, dir=sample_dir, prefix="tmp.inserts.", mode='w')
command.run("samtools view -f 66 " + file + " | cut -f 9 | sed 's/^-//'", tempfile.name)
insert_count = 0
insert_sum = 0
with open(tempfile.name) as f:
for line in f:
try:
insert_sum += int(line)
insert_count += 1
except ValueError:
pass
os.unlink(tempfile.name)
if insert_count > 0 and insert_sum > 0:
ave_insert_size = float(insert_sum) / float(insert_count)
ave_insert_size = "%.2f" % ave_insert_size
else:
handle_error("Cannot calculate mean insert size.")
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Calculate mean depth from pileup file"))
#-------------------------
ave_pileup_depth = ""
file = os.path.join(sample_dir, "reads.all.pileup")
if verify_input_file("Pileup file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
ave_pileup_depth = metrics.get("avePileupDepth", "") # reuse already fresh metrics
if ave_pileup_depth:
verbose_print("Reusing previously calculated mean pileup depth")
else:
depth_sum = 0
with open(file) as f:
for line in f:
tokens = line.split()
try:
depth_sum += int(tokens[3])
except (ValueError, IndexError):
pass
reference_length = 0
for record in SeqIO.parse(reference_file_path, "fasta"):
reference_length += len(record)
if depth_sum > 0 and reference_length > 0:
#print("depth_sum=%i" % depth_sum);
#print("reference_length=%i" % reference_length)
ave_pileup_depth = float(depth_sum) / float(reference_length)
ave_pileup_depth = "%.2f" % ave_pileup_depth
else:
handle_error("Cannot calculate mean pileup depth.")
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Count number of high confidence SNP positions from phase 1 vcf file"))
#-------------------------
phase1_snps = ""
excluded_sample = ""
file = os.path.join(sample_dir, "var.flt.vcf")
if verify_input_file("VCF file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase1_snps = metrics.get("phase1Snps", "") # reuse already fresh metrics
if phase1_snps:
verbose_print("Reusing previously calculated phase1 snps")
else:
phase1_snps = count_vcf_file_snps(file)
# Flag excessive snps
if max_allowed_snps > 0 and phase1_snps > max_allowed_snps:
excluded_sample = "Excluded"
handle_error("Excluded: exceeded %i maxsnps." % max_allowed_snps)
phase1_snps = str(phase1_snps)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Count number of filter_regions preserved high confidence SNP positions from phase 1 vcf file"))
#-------------------------
phase1_snps_preserved = ""
excluded_sample_preserved = ""
file = os.path.join(sample_dir, "var.flt_preserved.vcf")
if verify_input_file("VCF file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase1_snps_preserved = metrics.get("phase1SnpsPreserved", "") # reuse already fresh metrics
if phase1_snps_preserved:
verbose_print("Reusing previously calculated preserved phase1 snps")
else:
phase1_snps_preserved = count_vcf_file_snps(file)
# Flag excessive snps
if max_allowed_snps > 0 and phase1_snps_preserved > max_allowed_snps:
excluded_sample_preserved = "Excluded"
handle_error("Excluded: preserved exceeded %i maxsnps." % max_allowed_snps)
phase1_snps_preserved = str(phase1_snps_preserved)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Count number of consensus snps from consensus vcf file"))
#-------------------------
phase2_snps = ""
file = os.path.join(sample_dir, consensus_vcf_file_name)
if verify_input_file("Consensus VCF file", file):
# Omit the phase2 snp count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase2_snps = metrics.get("snps", "") # reuse already fresh metrics
if phase2_snps:
verbose_print("Reusing previously calculated phase2 snps")
else:
phase2_snps = count_vcf_file_snps(file)
phase2_snps = str(phase2_snps)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Count number of preserved consensus snps from consensus vcf file"))
#-------------------------
phase2_snps_preserved = ""
file = os.path.join(sample_dir, consensus_preserved_vcf_file_name)
if verify_input_file("Consensus VCF file", file):
# Omit the phase2 snp count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample_preserved != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase2_snps_preserved = metrics.get("snpsPreserved", "") # reuse already fresh metrics
if phase2_snps_preserved:
verbose_print("Reusing previously calculated preserved phase2 snps")
else:
phase2_snps_preserved = count_vcf_file_snps(file)
phase2_snps_preserved = str(phase2_snps_preserved)
#------------------------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Count missing positions in the snp matrix"))
#------------------------------------------
missing_pos = ""
file = os.path.join(sample_dir, consensus_fasta_file_name)
if verify_input_file("Consensus fasta file", file):
# Omit the phase2 gap count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
missing_pos = metrics.get("missingPos", "") # reuse already fresh metrics
if missing_pos:
verbose_print("Reusing previously calculated missing positions")
else:
missing_pos = count_missing_snp_matrix_positions(file, sample_id)
missing_pos = str(missing_pos)
#------------------------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Count missing positions in the preserved snp matrix"))
#------------------------------------------
missing_pos_preserved = ""
file = os.path.join(sample_dir, consensus_preserved_fasta_file_name)
if verify_input_file("Consensus fasta file", file):
# Omit the phase2 gap count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample_preserved != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
missing_pos_preserved = metrics.get("missingPosPreserved", "") # reuse already fresh metrics
if missing_pos_preserved:
verbose_print("Reusing previously calculated missing positions")
else:
missing_pos_preserved = count_missing_snp_matrix_positions(file, sample_id)
missing_pos_preserved = str(missing_pos_preserved)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Print results"))
#-------------------------
with open(metrics_file_path, "w") as f:
print("sample=" + '"' + sample_id + '"', file=f)
print("fastqFileList=" + '"' + fastq_file_list + '"', file=f)
print("fastqFileSize=" + str(fastq_file_size), file=f)
print("machine=" + machine, file=f)
print("flowcell=" + flowcell, file=f)
print("numberReads=" + num_reads, file=f)
print("numberDupReads=" + num_dup_reads, file=f)
print("percentReadsMapped=" + percent_reads_mapped, file=f)
print("aveInsertSize=" + ave_insert_size, file=f)
print("avePileupDepth=" + ave_pileup_depth, file=f)
print("phase1Snps=" + phase1_snps, file=f)
print("phase1SnpsPreserved=" + phase1_snps_preserved, file=f)
print("snps=" + phase2_snps, file=f)
print("snpsPreserved=" + phase2_snps_preserved, file=f)
print("missingPos=" + missing_pos, file=f)
print("missingPosPreserved=" + missing_pos_preserved, file=f)
print("excludedSample=" + excluded_sample, file=f)
print("excludedSamplePreserved=" + excluded_sample_preserved, file=f)
print("errorList=" + '"' + ' '.join(error_list) + '"', file=f)
def call_sites(args):
"""Find the sites with SNPs in a sample.
The sample alignment is sorted, duplicate reads are removed, a pileup is generated, and
snps are called.
This function expects, or creates '(*)', the following files arranged
in the following way:
reference
referenceFile.fasta
samples
sample_name_one/reads.sam
sample_name_one/reads.unsorted.bam*
sample_name_one/reads.sorted.bam*
sample_name_one/reads.sorted.deduped.bam*
sample_name_one/reads.all.pileup*
sample_name_one/var.flt.vcf*
The input files are created outside of this function. The package
documentation provides an example of preparing these files based on the
lambda_virus sequence that is used as one test for this package.
Parameters
----------
args : argparse.Namespace
referenceFile : File path of the reference fasta file
sampleDir : Relative or absolute directory of the sample
"""
utils.print_log_header(classpath=True)
utils.print_arguments(args)
#==========================================================================
# Validate inputs
#==========================================================================
# Verify reference fasta file exists and is not empty
reference_file_path = args.referenceFile
utils.verify_non_empty_input_files("Reference file", [reference_file_path], error_handler="global")
sample_dir = args.sampleDir
sam_file = os.path.join(sample_dir, "reads.sam")
utils.verify_non_empty_input_files("Sample SAM file", [sam_file], error_handler="sample")
sample_id = utils.sample_id_from_dir(sample_dir)
#==========================================================================
# Convert sam to bam file, selecting only the mapped reads
#==========================================================================
# Check for fresh bam file; if not, convert to bam file with only mapped reads
unsorted_bam_file = os.path.join(sample_dir, "reads.unsorted.bam")
needs_rebuild = utils.target_needs_rebuild([sam_file], unsorted_bam_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# Unsorted bam file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
version_str = utils.extract_version_str("SAMtools", "samtools 2>&1 > /dev/null")
# Substitute the default parameters if the user did not specify samtools view parameters
samtools_samfilter_params = os.environ.get("SamtoolsSamFilter_ExtraParams") or "-F 4"
command_line = "samtools view -S -b " + samtools_samfilter_params + " -o " + unsorted_bam_file + ' ' + sam_file
verbose_print("# Convert sam file to bam file with only mapped positions.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, sys.stdout)
utils.sample_error_on_missing_file(unsorted_bam_file, "samtools view")
verbose_print("")
#==========================================================================
# Sort the BAM file
#==========================================================================
# Check for fresh sorted bam file; if not, sort it
sorted_bam_file = os.path.join(sample_dir, "reads.sorted.bam")
needs_rebuild = utils.target_needs_rebuild([unsorted_bam_file], sorted_bam_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# Sorted bam file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
version_str = utils.extract_version_str("SAMtools", "samtools 2>&1 > /dev/null")
samtools_sort_extra_params = os.environ.get("SamtoolsSort_ExtraParams") or ""
# Inspect the samtools version to determine how to execute samtools
# Use the -o FILE command line option with SAMtools 1.3 and higher
samtools_version = version_str.split()[-1] # just the number
if samtools_version < "1.3":
command_line = "samtools sort " + samtools_sort_extra_params + ' ' + unsorted_bam_file + ' ' + os.path.join(sample_dir, "reads.sorted")
else:
command_line = "samtools sort " + samtools_sort_extra_params + " -o " + sorted_bam_file + ' ' + unsorted_bam_file
verbose_print("# Convert bam to sorted bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, sys.stdout)
utils.sample_error_on_missing_file(sorted_bam_file, "samtools sort")
verbose_print("")
#==========================================================================
# Mark duplicate reads, so they will be ignored in subsequent steps
#==========================================================================
remove_duplicate_reads = os.environ.get("RemoveDuplicateReads") or "true"
remove_duplicate_reads = remove_duplicate_reads.lower()
if remove_duplicate_reads == "true":
# Check for fresh deduped bam file; if not, remove duplicate reads
deduped_bam_file = os.path.join(sample_dir, "reads.sorted.deduped.bam")
needs_rebuild = utils.target_needs_rebuild([sorted_bam_file], deduped_bam_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# Deduped bam file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
classpath = os.environ.get("CLASSPATH")
if not classpath or "picard" not in classpath.lower():
utils.global_error("Error: cannot execute Picard. Define the path to Picard in the CLASSPATH environment variable.")
else:
version_str = utils.extract_version_str("Picard", "java picard.cmdline.PicardCommandLine MarkDuplicates --version 2>&1")
picard_jvm_extra_params = os.environ.get("PicardJvm_ExtraParams") or ""
picard_mark_duplicates_extra_params = os.environ.get("PicardMarkDuplicates_ExtraParams") or ""
tmpdir = os.environ.get("TMPDIR") or os.environ.get("TMP_DIR")
tmp_option = " TMP_DIR=" + tmpdir if tmpdir else ""
command_line = "java " + picard_jvm_extra_params + ' ' + "picard.cmdline.PicardCommandLine MarkDuplicates INPUT=" + sorted_bam_file + " OUTPUT=" + deduped_bam_file + " METRICS_FILE=" + os.path.join(sample_dir, "duplicate_reads_metrics.txt") + tmp_option + ' ' + picard_mark_duplicates_extra_params
verbose_print("# Remove duplicate reads from bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, sys.stdout)
utils.sample_error_on_missing_file(deduped_bam_file, "picard MarkDuplicates")
verbose_print("")
pileup_input_file = deduped_bam_file
else:
pileup_input_file = sorted_bam_file
#==========================================================================
# Create the pileup file
#==========================================================================
# Check for fresh pileup; if not, create it
pileup_file = os.path.join(sample_dir, "reads.all.pileup")
needs_rebuild = utils.target_needs_rebuild([pileup_input_file, reference_file_path], pileup_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# Pileup file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
version_str = utils.extract_version_str("SAMtools", "samtools 2>&1 > /dev/null")
samtools_mpileup_extra_params = os.environ.get("SamtoolsMpileup_ExtraParams") or ""
command_line = "samtools mpileup " + samtools_mpileup_extra_params + " -f " + reference_file_path + ' ' + pileup_input_file
verbose_print("# Create pileup from bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, pileup_file)
utils.sample_error_on_missing_file(pileup_file, "samtools mpileup")
verbose_print("")
#==========================================================================
# Find the sites with SNPs
#==========================================================================
# Check for fresh unfiltered vcf; if not, create it
vcf_file = os.path.join(sample_dir, "var.flt.vcf")
needs_rebuild = utils.target_needs_rebuild([pileup_file], vcf_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# VCF file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
classpath = os.environ.get("CLASSPATH")
if not classpath or "varscan" not in classpath.lower():
utils.global_error("Error: cannot execute VarScan. Define the path to VarScan in the CLASSPATH environment variable.")
else:
version_str = utils.extract_version_str("VarScan", "java net.sf.varscan.VarScan 2>&1 > /dev/null | head -n 1 | cut -d ' ' -f 2")
varscan_jvm_extra_params = os.environ.get("VarscanJvm_ExtraParams") or ""
varscan_mpileup2snp_extra_params = os.environ.get("VarscanMpileup2snp_ExtraParams") or ""
command_line = "java " + varscan_jvm_extra_params + " net.sf.varscan.VarScan mpileup2snp " + pileup_file + " --output-vcf 1 " + varscan_mpileup2snp_extra_params
verbose_print("# Create vcf file")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, vcf_file)
utils.sample_error_on_missing_file(vcf_file, "VarScan")
utils.sample_error_on_file_contains(vcf_file, "OutOfMemoryError", "VarScan")
utils.sample_error_on_file_contains(vcf_file, "Insufficient", "VarScan")
def collect_metrics(args):
"""Collect the quality metrics and SNP metrics for a sample.
This function expects, or creates '(*)', the following files arranged
in the following way:
reference
referenceFile.fasta
samples
sample_name_one/*.fastq.gz
sample_name_one/reads.sam
sample_name_one/reads.sorted.deduped.bam
sample_name_one/reads.sorted.bam
sample_name_one/reads.all.pileup
sample_name_one/var.flt.vcf
sample_name_one/var.flt_preserved.vcf
sample_name_one/consensus.fasta
sample_name_one/consensus_preserved.fasta
sample_name_one/consensus.vcf
sample_name_one/consensus_preserved.vcf
sample_name_one/metrics*
The input files are created outside of this function. The package
documentation provides an example of preparing these files based on the
lambda_virus sequence that is used as one test for this package.
Parameters
----------
args : argparse.Namespace
referenceFile : File path of the reference fasta file
sampleDir : Relative or absolute directory of the sample
consensusFastaFileName : File name of the consensus fasta file which must exist in the sample directory
consensusPreservedFastaFileName : File name of the consensus preserved fasta file which must exist in the sample directory
consensusVcfFileName : File name of the consensus vcf file which must exist in the sample directory
consensusPreservedVcfFileName : File name of the consensus preserved vcf file which must exist in the sample directory
maxSnps : Maximum allowed number of SNPs per sample
metricsFile : Output file. Relative or absolute path to the metrics file
"""
utils.print_log_header(classpath=True)
utils.print_arguments(args)
#==========================================================================
# Validate inputs
#==========================================================================
# Verify reference fasta file exists and is not empty
reference_file_path = args.referenceFile
utils.verify_non_empty_input_files("Reference file", [reference_file_path],
error_handler="global")
sample_dir = args.sampleDir
utils.verify_non_empty_directory("Sample directory",
sample_dir,
error_handler="sample",
continue_possible=False)
metrics_file_path = args.metricsFile
max_allowed_snps = args.maxSnps
consensus_vcf_file_name = args.consensusVcfFileName
consensus_preserved_vcf_file_name = args.consensusPreservedVcfFileName
consensus_fasta_file_name = args.consensusFastaFileName
consensus_preserved_fasta_file_name = args.consensusPreservedFastaFileName
sample_id = utils.sample_id_from_dir(sample_dir)
#==========================================================================
# Read existing metrics file so some metrics can be reused
#==========================================================================
try:
metrics = utils.read_properties(metrics_file_path)
except IOError:
metrics = dict()
#-------------------------
verbose_print(
"# %s %s" %
(utils.timestamp(), "Get machine and flowcell from fastq header"))
#-------------------------
machine = ""
flowcell = ""
fastq_files = fastq.list_fastq_files(sample_dir)
fastq_files = [f for f in fastq_files
if os.path.isfile(f)] # Exclude broken symlinks
if not fastq_files:
handle_error("No fastq files were found.")
else:
tags = fastq.extract_metadata_tags(fastq_files[0])
if tags:
machine = tags.instrument or ""
flowcell = tags.flow_cell or ""
#-------------------------
verbose_print("# %s %s" %
(utils.timestamp(), "Sum file sizes of paired fastq files"))
#-------------------------
fastq_file_size = ""
fastq_file_list = ""
if fastq_files:
fastq_file_size = sum([os.path.getsize(file) for file in fastq_files])
# Make a comma separated list of just the fastq file names without directories
fastq_file_list = [os.path.basename(file) for file in fastq_files]
fastq_file_list = ", ".join(fastq_file_list)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(
), "Calculate number of reads, %mapped, %proper pair, and ave insert size from sam file"
))
#-------------------------
num_reads = ""
percent_reads_mapped = ""
percent_proper_pair = ""
ave_insert_size = ""
file = os.path.join(sample_dir, "reads.sam")
if verify_input_file("SAM file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
num_reads = metrics.get("numberReads",
"") # reuse already fresh metrics
percent_reads_mapped = metrics.get(
"percentReadsMapped", "") # reuse already fresh metrics
percent_proper_pair = metrics.get(
"percentProperPair", "") # reuse already fresh metrics
ave_insert_size = metrics.get("aveInsertSize",
"") # reuse already fresh metrics
missing_any_metrics = not all([
num_reads, percent_reads_mapped, percent_proper_pair,
ave_insert_size
])
if not missing_any_metrics:
verbose_print(
"Reusing previously calculated number of reads, %mapped, %proper pair, and ave insert size"
)
else:
tempfile_path = os.path.join(sample_dir, "tmp.sam.stats")
try:
command.run("samtools stats " + file, tempfile_path)
except subprocess.CalledProcessError:
pass # the error message has already been printed to stderr
with open(tempfile_path) as f:
for line in f:
lower_line = line.lower()
split_line = line.strip().split('\t')
if "raw total sequences:" in lower_line:
num_reads = split_line[2]
continue
if "reads mapped:" in lower_line:
reads_mapped = split_line[2]
try:
percent_reads_mapped = 100.0 * float(
reads_mapped) / float(num_reads)
percent_reads_mapped = "%.2f" % percent_reads_mapped
except ValueError:
percent_reads_mapped = ""
continue
if "reads properly paired:" in lower_line:
proper_pairs = split_line[2]
try:
percent_proper_pair = 100.0 * float(
proper_pairs) / float(num_reads)
percent_proper_pair = "%.2f" % percent_proper_pair
except ValueError:
percent_proper_pair = ""
continue
if "insert size average:" in lower_line:
ave_insert_size = split_line[2]
continue
os.unlink(tempfile_path)
missing_any_metrics = not all([
num_reads, percent_reads_mapped, percent_proper_pair,
ave_insert_size
])
if missing_any_metrics:
missing_list = []
if not num_reads:
missing_list.append("number of reads")
if not percent_reads_mapped:
missing_list.append("percent reads mapped")
if not percent_proper_pair:
missing_list.append("percent proper pair")
if not ave_insert_size:
missing_list.append("ave insert size")
error_text = "Cannot calculate " + ", ".join(
missing_list) + '.'
handle_error(error_text)
#-------------------------
# Calculate number of duplicate reads from deduped bam file
#-------------------------
num_dup_reads = ""
remove_duplicate_reads = os.environ.get("RemoveDuplicateReads") or "true"
remove_duplicate_reads = remove_duplicate_reads.lower()
if remove_duplicate_reads == "true":
verbose_print(
"# %s %s" %
(utils.timestamp(),
"Calculate number of duplicate reads from deduped bam file"))
file = os.path.join(sample_dir, "reads.sorted.deduped.bam")
if verify_input_file("Deduped BAM file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file],
metrics_file_path)
if not args.forceFlag and not needs_rebuild:
num_dup_reads = metrics.get("numberDupReads",
"") # reuse already fresh metrics
if num_dup_reads:
verbose_print(
"Reusing previously calculated number of duplicate reads")
else:
num_dup_reads = command.run("samtools view -S -c -f 1024 " +
file)
num_dup_reads = num_dup_reads.strip()
#-------------------------
verbose_print("# %s %s" %
(utils.timestamp(), "Calculate mean depth from pileup file"))
#-------------------------
ave_pileup_depth = ""
file = os.path.join(sample_dir, "reads.all.pileup")
if verify_input_file("Pileup file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
ave_pileup_depth = metrics.get("avePileupDepth",
"") # reuse already fresh metrics
if ave_pileup_depth:
verbose_print("Reusing previously calculated mean pileup depth")
else:
depth_sum = 0
with open(file) as f:
for line in f:
tokens = line.split()
try:
depth_sum += int(tokens[3])
except (ValueError, IndexError):
pass
reference_length = 0
for record in SeqIO.parse(reference_file_path, "fasta"):
reference_length += len(record)
if depth_sum > 0 and reference_length > 0:
#print("depth_sum=%i" % depth_sum);
#print("reference_length=%i" % reference_length)
ave_pileup_depth = float(depth_sum) / float(reference_length)
ave_pileup_depth = "%.2f" % ave_pileup_depth
else:
handle_error("Cannot calculate mean pileup depth.")
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(
), "Count number of high confidence SNP positions from phase 1 vcf file"))
#-------------------------
phase1_snps = ""
excluded_sample = ""
file = os.path.join(sample_dir, "var.flt.vcf")
if verify_input_file("VCF file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase1_snps = metrics.get("phase1Snps",
"") # reuse already fresh metrics
if phase1_snps:
verbose_print("Reusing previously calculated phase1 snps")
else:
phase1_snps = count_vcf_file_snps(file)
# Flag excessive snps
if max_allowed_snps > 0 and phase1_snps > max_allowed_snps:
excluded_sample = "Excluded"
handle_error("Excluded: exceeded %i maxsnps." % max_allowed_snps)
phase1_snps = str(phase1_snps)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(
), "Count number of filter_regions preserved high confidence SNP positions from phase 1 vcf file"
))
#-------------------------
phase1_snps_preserved = ""
excluded_sample_preserved = ""
file = os.path.join(sample_dir, "var.flt_preserved.vcf")
if verify_input_file("VCF file", file):
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file], metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase1_snps_preserved = metrics.get(
"phase1SnpsPreserved", "") # reuse already fresh metrics
if phase1_snps_preserved:
verbose_print(
"Reusing previously calculated preserved phase1 snps")
else:
phase1_snps_preserved = count_vcf_file_snps(file)
# Flag excessive snps
if max_allowed_snps > 0 and phase1_snps_preserved > max_allowed_snps:
excluded_sample_preserved = "Excluded"
handle_error("Excluded: preserved exceeded %i maxsnps." %
max_allowed_snps)
phase1_snps_preserved = str(phase1_snps_preserved)
#-------------------------
verbose_print("# %s %s" %
(utils.timestamp(),
"Count number of consensus snps from consensus vcf file"))
#-------------------------
phase2_snps = ""
file = os.path.join(sample_dir, consensus_vcf_file_name)
if verify_input_file("Consensus VCF file", file):
# Omit the phase2 snp count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file],
metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase2_snps = metrics.get("snps",
"") # reuse already fresh metrics
if phase2_snps:
verbose_print("Reusing previously calculated phase2 snps")
else:
phase2_snps = count_vcf_file_snps(file)
phase2_snps = str(phase2_snps)
#-------------------------
verbose_print(
"# %s %s" %
(utils.timestamp(),
"Count number of preserved consensus snps from consensus vcf file"))
#-------------------------
phase2_snps_preserved = ""
file = os.path.join(sample_dir, consensus_preserved_vcf_file_name)
if verify_input_file("Consensus VCF file", file):
# Omit the phase2 snp count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample_preserved != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file],
metrics_file_path)
if not args.forceFlag and not needs_rebuild:
phase2_snps_preserved = metrics.get(
"snpsPreserved", "") # reuse already fresh metrics
if phase2_snps_preserved:
verbose_print(
"Reusing previously calculated preserved phase2 snps")
else:
phase2_snps_preserved = count_vcf_file_snps(file)
phase2_snps_preserved = str(phase2_snps_preserved)
#------------------------------------------
verbose_print(
"# %s %s" %
(utils.timestamp(), "Count missing positions in the snp matrix"))
#------------------------------------------
missing_pos = ""
file = os.path.join(sample_dir, consensus_fasta_file_name)
if verify_input_file("Consensus fasta file", file):
# Omit the phase2 gap count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file],
metrics_file_path)
if not args.forceFlag and not needs_rebuild:
missing_pos = metrics.get("missingPos",
"") # reuse already fresh metrics
if missing_pos:
verbose_print(
"Reusing previously calculated missing positions")
else:
missing_pos = count_missing_snp_matrix_positions(
file, sample_id)
missing_pos = str(missing_pos)
#------------------------------------------
verbose_print("# %s %s" %
(utils.timestamp(),
"Count missing positions in the preserved snp matrix"))
#------------------------------------------
missing_pos_preserved = ""
file = os.path.join(sample_dir, consensus_preserved_fasta_file_name)
if verify_input_file("Consensus fasta file", file):
# Omit the phase2 gap count if the sample is excluded.
# It will be meaningless since this sample's phase1 snps are excluded from the snplist.
if excluded_sample_preserved != "Excluded":
# Metrics already freshly collected?
needs_rebuild = utils.target_needs_rebuild([file],
metrics_file_path)
if not args.forceFlag and not needs_rebuild:
missing_pos_preserved = metrics.get(
"missingPosPreserved", "") # reuse already fresh metrics
if missing_pos_preserved:
verbose_print(
"Reusing previously calculated missing positions")
else:
missing_pos_preserved = count_missing_snp_matrix_positions(
file, sample_id)
missing_pos_preserved = str(missing_pos_preserved)
#-------------------------
verbose_print("# %s %s" % (utils.timestamp(), "Print results"))
#-------------------------
with open(metrics_file_path, "w") as f:
print("sample=" + '"' + sample_id + '"', file=f)
print("fastqFileList=" + '"' + fastq_file_list + '"', file=f)
print("fastqFileSize=" + str(fastq_file_size), file=f)
print("machine=" + machine, file=f)
print("flowcell=" + flowcell, file=f)
print("numberReads=" + num_reads, file=f)
print("numberDupReads=" + num_dup_reads, file=f)
print("percentReadsMapped=" + percent_reads_mapped, file=f)
print("percentProperPair=" + percent_proper_pair, file=f)
print("aveInsertSize=" + ave_insert_size, file=f)
print("avePileupDepth=" + ave_pileup_depth, file=f)
print("phase1Snps=" + phase1_snps, file=f)
print("phase1SnpsPreserved=" + phase1_snps_preserved, file=f)
print("snps=" + phase2_snps, file=f)
print("snpsPreserved=" + phase2_snps_preserved, file=f)
print("missingPos=" + missing_pos, file=f)
print("missingPosPreserved=" + missing_pos_preserved, file=f)
print("excludedSample=" + excluded_sample, file=f)
print("excludedSamplePreserved=" + excluded_sample_preserved, file=f)
print("errorList=" + '"' + ' '.join(error_list) + '"', file=f)
def call_sites(args):
"""Find the sites with SNPs in a sample.
The sample alignment is sorted, duplicate reads are removed, a pileup is generated, and
snps are called.
This function expects, or creates '(*)', the following files arranged
in the following way:
reference
referenceFile.fasta
samples
sample_name_one/reads.sorted.deduped.indelrealigned.bam
sample_name_one/reads.all.pileup*
sample_name_one/var.flt.vcf*
The input files are created outside of this function. The package
documentation provides an example of preparing these files based on the
lambda_virus sequence that is used as one test for this package.
Parameters
----------
args : argparse.Namespace
referenceFile : File path of the reference fasta file
sampleDir : Relative or absolute directory of the sample
"""
utils.print_log_header(classpath=True)
utils.print_arguments(args)
#==========================================================================
# Validate inputs
#==========================================================================
# Verify reference fasta file exists and is not empty
reference_file_path = args.referenceFile
utils.verify_non_empty_input_files("Reference file", [reference_file_path], error_handler="global")
sample_dir = args.sampleDir
remove_duplicate_reads = os.environ.get("RemoveDuplicateReads", "true").lower() == "true"
enable_local_realignment = os.environ.get("EnableLocalRealignment", "true").lower() == "true"
input_bam_file = os.path.join(sample_dir, "reads.sorted.bam")
input_bam_file = utils.add_file_suffix(input_bam_file, ".deduped", enable=remove_duplicate_reads)
input_bam_file = utils.add_file_suffix(input_bam_file, ".indelrealigned", enable=enable_local_realignment)
utils.verify_non_empty_input_files("Sample BAM file", [input_bam_file], error_handler="sample")
sample_id = utils.sample_id_from_dir(sample_dir)
#==========================================================================
# Create the pileup file
#==========================================================================
# Check for fresh pileup; if not, create it
pileup_file = os.path.join(sample_dir, "reads.all.pileup")
needs_rebuild = utils.target_needs_rebuild([input_bam_file, reference_file_path], pileup_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# Pileup file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
version_str = utils.extract_version_str("SAMtools", "samtools 2>&1 > /dev/null")
samtools_mpileup_extra_params = os.environ.get("SamtoolsMpileup_ExtraParams") or ""
command_line = "samtools mpileup " + samtools_mpileup_extra_params + " -f " + reference_file_path + ' ' + input_bam_file
verbose_print("# Create pileup from bam file.")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, pileup_file)
utils.sample_error_on_missing_file(pileup_file, "samtools mpileup")
verbose_print("")
#==========================================================================
# Find the sites with SNPs
#==========================================================================
# Check for fresh unfiltered vcf; if not, create it
vcf_file = os.path.join(sample_dir, "var.flt.vcf")
needs_rebuild = utils.target_needs_rebuild([pileup_file], vcf_file)
if not args.forceFlag and not needs_rebuild:
verbose_print("# VCF file is already freshly created for %s. Use the -f option to force a rebuild." % sample_id)
else:
jar_file_path = utils.find_path_in_path_list("VarScan", "CLASSPATH")
if not jar_file_path:
utils.global_error("Error: cannot execute VarScan. Define the path to VarScan.jar in the CLASSPATH environment variable.")
else:
version_str = utils.extract_version_str("VarScan", "java -jar " + jar_file_path + " 2>&1 > /dev/null | head -n 1 | cut -d ' ' -f 2")
varscan_jvm_extra_params = os.environ.get("VarscanJvm_ExtraParams") or ""
varscan_mpileup2snp_extra_params = os.environ.get("VarscanMpileup2snp_ExtraParams") or ""
command_line = "java " + varscan_jvm_extra_params + " -jar " + jar_file_path + " mpileup2snp " + pileup_file + " --output-vcf 1 " + varscan_mpileup2snp_extra_params
verbose_print("# Create vcf file")
verbose_print("# %s %s" % (utils.timestamp(), command_line))
verbose_print("# %s" % version_str)
command.run(command_line, vcf_file)
utils.sample_error_on_missing_file(vcf_file, "VarScan")
utils.sample_error_on_file_contains(vcf_file, "OutOfMemoryError", "VarScan")
utils.sample_error_on_file_contains(vcf_file, "Insufficient", "VarScan")