def align_reference(ref_fpath, output_dir, using_reads='all', calculate_coverage=False):
required_files = []
ref_name = qutils.name_from_fpath(ref_fpath)
cov_fpath = qconfig.cov_fpath or join(output_dir, ref_name + '.cov')
uncovered_fpath = add_suffix(cov_fpath, 'uncovered')
if using_reads != 'all':
cov_fpath = add_suffix(cov_fpath, using_reads)
uncovered_fpath = add_suffix(uncovered_fpath, using_reads)
insert_size_fpath = join(output_dir, ref_name + '.is.txt')
if not is_non_empty_file(uncovered_fpath):
required_files.append(uncovered_fpath)
if not is_non_empty_file(insert_size_fpath) and (using_reads == 'all' or using_reads == 'pe'):
required_files.append(insert_size_fpath)
temp_output_dir = join(output_dir, 'temp_output')
if not isdir(temp_output_dir):
os.makedirs(temp_output_dir)
log_path = join(output_dir, 'reads_stats.log')
err_fpath = join(output_dir, 'reads_stats.err')
correct_chr_names, sam_fpath, bam_fpath = align_single_file(ref_fpath, output_dir, temp_output_dir, log_path, err_fpath,
qconfig.max_threads, sam_fpath=qconfig.reference_sam,
bam_fpath=qconfig.reference_bam, required_files=required_files,
is_reference=True, alignment_only=True, using_reads=using_reads)
if not qconfig.optimal_assembly_insert_size or qconfig.optimal_assembly_insert_size == 'auto':
if using_reads == 'pe' and sam_fpath:
insert_size, std_dev = calculate_insert_size(sam_fpath, output_dir, ref_name)
if not insert_size:
logger.info(' Failed calculating insert size.')
else:
qconfig.optimal_assembly_insert_size = insert_size
elif using_reads == 'all' and is_non_empty_file(insert_size_fpath):
try:
insert_size = int(open(insert_size_fpath).readline())
if insert_size:
qconfig.optimal_assembly_insert_size = insert_size
except:
pass
if not required_files:
return sam_fpath, bam_fpath, uncovered_fpath
if not sam_fpath:
logger.info(' Failed detecting uncovered regions.')
return None, None
if calculate_coverage:
bam_mapped_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_fpath, 'mapped'))
bam_sorted_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_mapped_fpath, 'sorted'))
if is_non_empty_file(bam_sorted_fpath):
logger.info(' Using existing sorted BAM-file: ' + bam_sorted_fpath)
else:
sambamba_view(bam_fpath, bam_mapped_fpath, qconfig.max_threads, err_fpath, logger, filter_rule='not unmapped')
sort_bam(bam_mapped_fpath, bam_sorted_fpath, err_fpath, logger)
if not is_non_empty_file(uncovered_fpath) and calculate_coverage:
get_coverage(temp_output_dir, ref_fpath, ref_name, bam_fpath, bam_sorted_fpath, log_path, err_fpath,
correct_chr_names, cov_fpath, uncovered_fpath=uncovered_fpath, create_cov_files=False)
return sam_fpath, bam_fpath, uncovered_fpath
def align_reference(ref_fpath, output_dir, using_reads='all'):
required_files = []
ref_name = qutils.name_from_fpath(ref_fpath)
cov_fpath = qconfig.cov_fpath or join(output_dir, ref_name + '.cov')
uncovered_fpath = add_suffix(cov_fpath, 'uncovered')
if using_reads != 'all':
cov_fpath = add_suffix(cov_fpath, using_reads)
uncovered_fpath = add_suffix(uncovered_fpath, using_reads)
insert_size_fpath = join(output_dir, ref_name + '.is.txt')
if not is_non_empty_file(uncovered_fpath):
required_files.append(uncovered_fpath)
if not is_non_empty_file(insert_size_fpath) and (using_reads == 'all' or using_reads == 'paired_end'):
required_files.append(insert_size_fpath)
temp_output_dir = join(output_dir, 'temp_output')
if not isdir(temp_output_dir):
os.makedirs(temp_output_dir)
log_path = join(output_dir, 'reads_stats.log')
err_fpath = join(output_dir, 'reads_stats.err')
correct_chr_names, sam_fpath, bam_fpath = align_single_file(ref_fpath, output_dir, temp_output_dir, log_path, err_fpath,
qconfig.max_threads, sam_fpath=qconfig.reference_sam,
bam_fpath=qconfig.reference_bam, required_files=required_files,
is_reference=True, alignment_only=True, using_reads=using_reads)
qconfig.reference_sam = sam_fpath
qconfig.reference_bam = bam_fpath
if not qconfig.ideal_assembly_insert_size or qconfig.ideal_assembly_insert_size == 'auto':
if using_reads == 'paired_end' and sam_fpath:
insert_size = calculate_insert_size(sam_fpath, output_dir, ref_name)
if not insert_size:
logger.info(' Failed calculating insert size.')
else:
qconfig.ideal_assembly_insert_size = insert_size
if not required_files:
return bam_fpath, uncovered_fpath
if not sam_fpath:
logger.info(' Failed detecting uncovered regions.')
return None, None
bam_mapped_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_fpath, 'mapped'))
bam_sorted_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_mapped_fpath, 'sorted'))
if is_non_empty_file(bam_sorted_fpath):
logger.info(' Using existing sorted BAM-file: ' + bam_sorted_fpath)
else:
sambamba_view(bam_fpath, bam_mapped_fpath, qconfig.max_threads, err_fpath, logger, filter_rule='not unmapped')
sort_bam(bam_mapped_fpath, bam_sorted_fpath, err_fpath, logger)
if not is_non_empty_file(uncovered_fpath):
get_coverage(temp_output_dir, ref_fpath, ref_name, bam_fpath, bam_sorted_fpath, log_path, err_fpath,
correct_chr_names, cov_fpath, uncovered_fpath=uncovered_fpath, create_cov_files=False)
return bam_fpath, uncovered_fpath
def merge_sam_files(tmp_sam_fpaths, sam_fpath, bam_fpath, max_threads, err_fpath):
tmp_bam_fpaths = []
for tmp_sam_fpath in tmp_sam_fpaths:
if is_non_empty_file(tmp_sam_fpath):
tmp_bam_fpath = tmp_sam_fpath.replace('.sam', '.bam')
tmp_bam_sorted_fpath = add_suffix(tmp_bam_fpath, 'sorted')
if not is_non_empty_file(tmp_bam_sorted_fpath):
sort_bam(tmp_bam_fpath, tmp_bam_sorted_fpath, err_fpath, logger)
tmp_bam_fpaths.append(tmp_bam_sorted_fpath)
qutils.call_subprocess([sambamba_fpath('sambamba'), 'merge', '-t', str(max_threads), bam_fpath] + tmp_bam_fpaths,
stderr=open(err_fpath, 'a'), logger=logger)
sambamba_view(bam_fpath, sam_fpath, max_threads, err_fpath, logger)
return sam_fpath
def get_physical_coverage(output_dirpath, ref_name, bam_fpath, log_path, err_fpath, cov_fpath, chr_len_fpath):
if not isfile(bedtools_fpath('bamToBed')):
logger.info(' Failed calculating physical coverage...')
return None
raw_cov_fpath = add_suffix(cov_fpath, 'raw')
if not is_non_empty_file(raw_cov_fpath):
logger.info(' Calculating physical coverage...')
## keep properly mapped, unique, and non-duplicate read pairs only
bam_filtered_fpath = join(output_dirpath, ref_name + '.filtered.bam')
sambamba_view(bam_fpath, bam_filtered_fpath, qconfig.max_threads, err_fpath, logger,
filter_rule='proper_pair and not supplementary and not duplicate')
## sort by read names
bam_filtered_sorted_fpath = join(output_dirpath, ref_name + '.filtered.sorted.bam')
sort_bam(bam_filtered_fpath, bam_filtered_sorted_fpath, err_fpath, logger, sort_rule='-n')
bed_fpath = bam_to_bed(output_dirpath, ref_name, bam_filtered_sorted_fpath, err_fpath, logger, bedpe=True)
calculate_genome_cov(bed_fpath, raw_cov_fpath, chr_len_fpath, err_fpath, logger)
return raw_cov_fpath
def get_physical_coverage(output_dirpath, ref_name, bam_fpath, log_path, err_fpath, cov_fpath, chr_len_fpath):
if not isfile(bedtools_fpath('bamToBed')):
logger.info(' Failed calculating physical coverage...')
return None
raw_cov_fpath = add_suffix(cov_fpath, 'raw')
if not is_non_empty_file(raw_cov_fpath):
logger.info(' Calculating physical coverage...')
## keep properly mapped, unique, and non-duplicate read pairs only
bam_filtered_fpath = join(output_dirpath, ref_name + '.filtered.bam')
sambamba_view(bam_fpath, bam_filtered_fpath, qconfig.max_threads, err_fpath, logger,
filter_rule='proper_pair and not supplementary and not duplicate')
## sort by read names
bam_filtered_sorted_fpath = join(output_dirpath, ref_name + '.filtered.sorted.bam')
sort_bam(bam_filtered_fpath, bam_filtered_sorted_fpath, err_fpath, logger, sort_rule='-n')
bed_fpath = bam_to_bed(output_dirpath, ref_name, bam_filtered_sorted_fpath, err_fpath, logger, bedpe=True)
calculate_genome_cov(bed_fpath, raw_cov_fpath, chr_len_fpath, err_fpath, logger)
return raw_cov_fpath
def process_one_ref(cur_ref_fpath, output_dirpath, err_fpath, max_threads, bam_fpath=None, bed_fpath=None):
ref_name = qutils.name_from_fpath(cur_ref_fpath)
if not bam_fpath:
sam_fpath = join(output_dirpath, ref_name + '.sam')
bam_fpath = join(output_dirpath, ref_name + '.bam')
bam_sorted_fpath = join(output_dirpath, ref_name + '.sorted.bam')
else:
sam_fpath = bam_fpath.replace('.bam', '.sam')
bam_sorted_fpath = add_suffix(bam_fpath, 'sorted')
bed_fpath = bed_fpath or join(output_dirpath, ref_name + '.bed')
if is_non_empty_file(bed_fpath):
logger.info(' Using existing BED-file: ' + bed_fpath)
return bed_fpath
if not isfile(bam_sorted_fpath):
sambamba_view(sam_fpath, bam_fpath, qconfig.max_threads, err_fpath, logger, filter_rule='not unmapped and proper_pair')
sort_bam(bam_fpath, bam_sorted_fpath, err_fpath, logger, threads=max_threads)
if not is_non_empty_file(bam_sorted_fpath + '.bai'):
qutils.call_subprocess([sambamba_fpath('sambamba'), 'index', bam_sorted_fpath],
stderr=open(err_fpath, 'a'), logger=logger)
create_fai_file(cur_ref_fpath)
vcf_output_dirpath = join(output_dirpath, ref_name + '_gridss')
vcf_fpath = join(vcf_output_dirpath, ref_name + '.vcf')
if not is_non_empty_file(vcf_fpath):
if isdir(vcf_output_dirpath):
shutil.rmtree(vcf_output_dirpath, ignore_errors=True)
os.makedirs(vcf_output_dirpath)
max_mem = get_gridss_memory()
env = os.environ.copy()
env["PATH"] += os.pathsep + bwa_dirpath
bwa_index(cur_ref_fpath, err_fpath, logger)
qutils.call_subprocess(['java', '-ea', '-Xmx' + str(max_mem) + 'g', '-Dsamjdk.create_index=true', '-Dsamjdk.use_async_io_read_samtools=true',
'-Dsamjdk.use_async_io_write_samtools=true', '-Dsamjdk.use_async_io_write_tribble=true',
'-cp', get_gridss_fpath(), 'gridss.CallVariants', 'I=' + bam_sorted_fpath, 'O=' + vcf_fpath,
'ASSEMBLY=' + join(vcf_output_dirpath, ref_name + '.gridss.bam'), 'REFERENCE_SEQUENCE=' + cur_ref_fpath,
'WORKER_THREADS=' + str(max_threads), 'WORKING_DIR=' + vcf_output_dirpath],
stderr=open(err_fpath, 'a'), logger=logger, env=env)
if is_non_empty_file(vcf_fpath):
raw_bed_fpath = add_suffix(bed_fpath, 'raw')
filtered_bed_fpath = add_suffix(bed_fpath, 'filtered')
qutils.call_subprocess(['java', '-cp', get_gridss_fpath(), 'au.edu.wehi.idsv.VcfBreakendToBedpe',
'I=' + vcf_fpath, 'O=' + raw_bed_fpath, 'OF=' + filtered_bed_fpath, 'R=' + cur_ref_fpath,
'INCLUDE_HEADER=TRUE'], stderr=open(err_fpath, 'a'), logger=logger)
reformat_bedpe(raw_bed_fpath, bed_fpath)
return bed_fpath
def process_one_ref(cur_ref_fpath, output_dirpath, err_fpath, max_threads, bam_fpath=None, bed_fpath=None):
ref_name = qutils.name_from_fpath(cur_ref_fpath)
if not bam_fpath:
sam_fpath = join(output_dirpath, ref_name + '.sam')
bam_fpath = join(output_dirpath, ref_name + '.bam')
bam_sorted_fpath = join(output_dirpath, ref_name + '.sorted.bam')
else:
sam_fpath = bam_fpath.replace('.bam', '.sam')
bam_sorted_fpath = add_suffix(bam_fpath, 'sorted')
bed_fpath = bed_fpath or join(output_dirpath, ref_name + '.bed')
if is_non_empty_file(bed_fpath):
logger.info(' Using existing BED-file: ' + bed_fpath)
return bed_fpath
if not isfile(bam_sorted_fpath):
sambamba_view(sam_fpath, bam_fpath, qconfig.max_threads, err_fpath, logger, filter_rule='not unmapped')
sort_bam(bam_fpath, bam_sorted_fpath, err_fpath, logger, threads=max_threads)
if not is_non_empty_file(bam_sorted_fpath + '.bai'):
qutils.call_subprocess([sambamba_fpath('sambamba'), 'index', bam_sorted_fpath],
stderr=open(err_fpath, 'a'), logger=logger)
create_fai_file(cur_ref_fpath)
vcf_output_dirpath = join(output_dirpath, ref_name + '_gridss')
vcf_fpath = join(vcf_output_dirpath, ref_name + '.vcf')
if not is_non_empty_file(vcf_fpath):
if isdir(vcf_output_dirpath):
shutil.rmtree(vcf_output_dirpath, ignore_errors=True)
os.makedirs(vcf_output_dirpath)
max_mem = get_gridss_memory()
env = os.environ.copy()
env["PATH"] += os.pathsep + bwa_dirpath
bwa_index(cur_ref_fpath, err_fpath, logger)
qutils.call_subprocess(['java', '-ea', '-Xmx' + str(max_mem) + 'g', '-Dsamjdk.create_index=true', '-Dsamjdk.use_async_io_read_samtools=true',
'-Dsamjdk.use_async_io_write_samtools=true', '-Dsamjdk.use_async_io_write_tribble=true',
'-cp', get_gridss_fpath(), 'gridss.CallVariants', 'I=' + bam_sorted_fpath, 'O=' + vcf_fpath,
'ASSEMBLY=' + join(vcf_output_dirpath, ref_name + '.gridss.bam'), 'REFERENCE_SEQUENCE=' + cur_ref_fpath,
'WORKER_THREADS=' + str(max_threads), 'WORKING_DIR=' + vcf_output_dirpath],
stderr=open(err_fpath, 'a'), logger=logger, env=env)
if is_non_empty_file(vcf_fpath):
raw_bed_fpath = add_suffix(bed_fpath, 'raw')
filtered_bed_fpath = add_suffix(bed_fpath, 'filtered')
qutils.call_subprocess(['java', '-cp', get_gridss_fpath(), 'au.edu.wehi.idsv.VcfBreakendToBedpe',
'I=' + vcf_fpath, 'O=' + raw_bed_fpath, 'OF=' + filtered_bed_fpath, 'R=' + cur_ref_fpath,
'INCLUDE_HEADER=TRUE'], stderr=open(err_fpath, 'a'), logger=logger)
reformat_bedpe(raw_bed_fpath, bed_fpath)
return bed_fpath
def merge_sam_files(tmp_sam_fpaths, sam_fpath, bam_fpath, output_dir, max_threads, err_fpath):
merged_bam_fpath = add_suffix(bam_fpath, 'merged')
tmp_bam_fpaths = []
for tmp_sam_fpath in tmp_sam_fpaths:
if is_non_empty_file(tmp_sam_fpath):
tmp_bam_fpath = tmp_sam_fpath.replace('.sam', '.bam')
tmp_bam_sorted_fpath = add_suffix(tmp_bam_fpath, 'sorted')
if not is_non_empty_file(tmp_bam_sorted_fpath):
sambamba_view(tmp_sam_fpath, tmp_bam_fpath, max_threads, err_fpath, logger, filter_rule=None)
sort_bam(tmp_bam_fpath, tmp_bam_sorted_fpath, err_fpath, logger)
tmp_bam_fpaths.append(tmp_bam_sorted_fpath)
qutils.call_subprocess([sambamba_fpath('sambamba'), 'merge', '-t', str(max_threads), merged_bam_fpath] + tmp_bam_fpaths,
stderr=open(err_fpath, 'a'), logger=logger)
qutils.call_subprocess([sambamba_fpath('sambamba'), 'markdup', '-r', '-t', str(max_threads), '--tmpdir',
output_dir, merged_bam_fpath, bam_fpath],
stderr=open(err_fpath, 'a'), logger=logger)
sambamba_view(bam_fpath, sam_fpath, max_threads, err_fpath, logger)
return merged_bam_fpath
def run_aligner(read_fpaths, ref_fpath, sam_fpath, out_sam_fpaths, output_dir, err_fpath, max_threads, reads_type):
bwa_cmd = bwa_fpath('bwa') + ' mem -t ' + str(max_threads)
insert_sizes = []
for idx, reads in enumerate(read_fpaths):
if isinstance(reads, str):
if reads_type == 'pacbio' or reads_type == 'nanopore':
if reads_type == 'pacbio':
preset = ' -ax map-pb '
else:
preset = ' -ax map-ont '
cmdline = minimap_fpath() + ' -t ' + str(max_threads) + preset + ref_fpath + ' ' + reads
else:
cmdline = bwa_cmd + (' -p ' if reads_type == 'pe' else ' ') + ref_fpath + ' ' + reads
else:
read1, read2 = reads
cmdline = bwa_cmd + ' ' + ref_fpath + ' ' + read1 + ' ' + read2
output_fpath = add_suffix(sam_fpath, reads_type + str(idx + 1))
bam_fpath = output_fpath.replace('.sam', '.bam')
if not is_non_empty_file(output_fpath):
qutils.call_subprocess(shlex.split(cmdline), stdout=open(output_fpath, 'w'), stderr=open(err_fpath, 'a'), logger=logger)
if not is_non_empty_file(bam_fpath):
if not is_non_empty_file(bam_fpath):
sambamba_view(output_fpath, bam_fpath, max_threads, err_fpath, logger, filter_rule=None)
if reads_type == 'pe':
bam_dedup_fpath = add_suffix(bam_fpath, 'dedup')
qutils.call_subprocess([sambamba_fpath('sambamba'), 'markdup', '-r', '-t', str(max_threads), '--tmpdir',
output_dir, bam_fpath, bam_dedup_fpath],
stderr=open(err_fpath, 'a'), logger=logger)
if exists(bam_dedup_fpath):
shutil.move(bam_dedup_fpath, bam_fpath)
if reads_type == 'pe':
insert_size, std_dev = calculate_insert_size(output_fpath, output_dir, basename(sam_fpath))
if insert_size < qconfig.optimal_assembly_max_IS:
insert_sizes.append(insert_size)
out_sam_fpaths.append(output_fpath)
if insert_sizes:
qconfig.optimal_assembly_insert_size = max(insert_sizes)
ref_name = qutils.name_from_fpath(ref_fpath)
insert_size_fpath = join(output_dir, '..', ref_name + '.is.txt')
with open(insert_size_fpath, 'w') as out:
out.write(str(qconfig.optimal_assembly_insert_size))
def get_joiners(ref_name, sam_fpath, bam_fpath, output_dirpath, err_fpath,
using_reads):
bam_filtered_fpath = add_suffix(bam_fpath, 'filtered')
if not is_non_empty_file(bam_filtered_fpath):
filter_rule = 'not unmapped and not supplementary and not secondary_alignment'
sambamba_view(bam_fpath,
bam_filtered_fpath,
qconfig.max_threads,
err_fpath,
logger,
filter_rule=filter_rule)
bam_sorted_fpath = add_suffix(bam_fpath, 'sorted')
if not is_non_empty_file(bam_sorted_fpath):
sort_bam(bam_filtered_fpath,
bam_sorted_fpath,
err_fpath,
logger,
sort_rule='-n')
bed_fpath = bam_to_bed(output_dirpath,
using_reads,
bam_sorted_fpath,
err_fpath,
logger,
bedpe=using_reads == 'mp')
intervals = defaultdict(list)
if using_reads == 'mp':
insert_size, std_dev = calculate_insert_size(sam_fpath,
output_dirpath,
ref_name,
reads_suffix='mp')
min_is = insert_size - std_dev
max_is = insert_size + std_dev
with open(bed_fpath) as bed:
for l in bed:
fs = l.split()
if using_reads == 'mp' and insert_size:
interval_len = int(fs[2]) - int(fs[1])
if min_is <= abs(interval_len) <= max_is:
intervals[fs[0]].append((int(fs[1]), int(fs[2])))
else:
intervals[fs[0]].append((int(fs[1]), int(fs[2])))
return intervals
def align_single_file(fpath, main_output_dir, output_dirpath, log_path, err_fpath, max_threads, sam_fpath=None, bam_fpath=None,
index=None, required_files=None, is_reference=False, alignment_only=False, using_reads='all'):
filename = qutils.name_from_fpath(fpath)
if not sam_fpath and bam_fpath:
sam_fpath = get_safe_fpath(output_dirpath, bam_fpath[:-4] + '.sam')
else:
sam_fpath = sam_fpath or join(output_dirpath, filename + '.sam')
bam_fpath = bam_fpath or get_safe_fpath(output_dirpath, sam_fpath[:-4] + '.bam')
if using_reads != 'all':
sam_fpath = join(output_dirpath, filename + '.' + using_reads + '.sam')
bam_fpath = sam_fpath.replace('.sam', '.bam')
if alignment_only or (is_reference and required_files and any(f.endswith('bed') for f in required_files)):
required_files.append(sam_fpath)
stats_fpath = get_safe_fpath(dirname(output_dirpath), filename + '.stat')
index_str = qutils.index_to_str(index) if index is not None else ''
reads_fpaths = qconfig.reads_fpaths
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
can_reuse = correct_chr_names is not None
if not can_reuse and not reads_fpaths:
return None, None, None
if correct_chr_names and (not required_files or all(isfile(fpath) for fpath in required_files)):
if not alignment_only:
if isfile(stats_fpath):
logger.info(' ' + index_str + 'Using existing flag statistics file ' + stats_fpath)
elif isfile(bam_fpath):
qutils.call_subprocess([sambamba_fpath('sambamba'), 'flagstat', '-t', str(max_threads), bam_fpath],
stdout=open(stats_fpath, 'w'), stderr=open(err_fpath, 'a'))
analyse_coverage(output_dirpath, fpath, correct_chr_names, bam_fpath, stats_fpath, err_fpath, logger)
if isfile(stats_fpath) or alignment_only:
return correct_chr_names, sam_fpath, bam_fpath
logger.info(' ' + index_str + 'Pre-processing reads...')
if is_non_empty_file(sam_fpath) and can_reuse:
logger.info(' ' + index_str + 'Using existing SAM-file: ' + sam_fpath)
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
elif is_non_empty_file(bam_fpath) and can_reuse:
logger.info(' ' + index_str + 'Using existing BAM-file: ' + bam_fpath)
sambamba_view(bam_fpath, sam_fpath, qconfig.max_threads, err_fpath, logger)
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
if (not correct_chr_names or not is_non_empty_file(sam_fpath)) and reads_fpaths:
if is_reference:
logger.info(' Running BWA for reference...')
else:
logger.info(' ' + index_str + 'Running BWA...')
# use absolute paths because we will change workdir
fpath = abspath(fpath)
sam_fpath = abspath(sam_fpath)
prev_dir = os.getcwd()
os.chdir(output_dirpath)
bwa_index(fpath, err_fpath, logger)
sam_fpaths = align_reads(fpath, sam_fpath, using_reads, main_output_dir, err_fpath, max_threads)
if len(sam_fpaths) > 1:
merge_sam_files(sam_fpaths, sam_fpath, bam_fpath, max_threads, err_fpath)
elif len(sam_fpaths) == 1:
shutil.move(sam_fpaths[0], sam_fpath)
tmp_bam_fpath = sam_fpaths[0].replace('.sam', '.bam')
if is_non_empty_file(tmp_bam_fpath):
shutil.move(tmp_bam_fpath, bam_fpath)
logger.info(' ' + index_str + 'Done.')
os.chdir(prev_dir)
if not is_non_empty_file(sam_fpath):
logger.error(' Failed running BWA for ' + fpath + '. See ' + log_path + ' for information.')
return None, None, None
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
elif not correct_chr_names or not is_non_empty_file(sam_fpath):
return None, None, None
if is_reference:
logger.info(' Sorting SAM-file for reference...')
else:
logger.info(' ' + index_str + 'Sorting SAM-file...')
if can_reuse and is_non_empty_file(bam_fpath) and all_read_names_correct(sam_fpath):
logger.info(' ' + index_str + 'Using existing BAM-file: ' + bam_fpath)
else:
correct_sam_fpath = join(output_dirpath, filename + '.' + using_reads + '.correct.sam') # write in output dir
sam_fpath = clean_read_names(sam_fpath, correct_sam_fpath)
sambamba_view(correct_sam_fpath, bam_fpath, max_threads, err_fpath, logger, filter_rule=None)
qutils.assert_file_exists(bam_fpath, 'bam file')
if not alignment_only:
if isfile(stats_fpath):
logger.info(' ' + index_str + 'Using existing flag statistics file ' + stats_fpath)
elif isfile(bam_fpath):
qutils.call_subprocess([sambamba_fpath('sambamba'), 'flagstat', '-t', str(max_threads), bam_fpath],
stdout=open(stats_fpath, 'w'), stderr=open(err_fpath, 'a'))
analyse_coverage(output_dirpath, fpath, correct_chr_names, bam_fpath, stats_fpath, err_fpath, logger)
if is_reference:
logger.info(' Analysis for reference is finished.')
else:
logger.info(' ' + index_str + 'Analysis is finished.')
return correct_chr_names, sam_fpath, bam_fpath
def run_processing_reads(contigs_fpaths, main_ref_fpath, meta_ref_fpaths, ref_labels, temp_output_dir, output_dir,
log_path, err_fpath):
required_files = []
bed_fpath, cov_fpath, physical_cov_fpath = None, None, None
if main_ref_fpath:
ref_name = qutils.name_from_fpath(main_ref_fpath)
bed_fpath = qconfig.bed or join(output_dir, ref_name + '.bed')
cov_fpath = qconfig.cov_fpath or join(output_dir, ref_name + '.cov')
physical_cov_fpath = qconfig.phys_cov_fpath or join(output_dir, ref_name + '.physical.cov')
required_files = [bed_fpath, cov_fpath, physical_cov_fpath]
if qconfig.no_sv:
logger.info(' Will not search Structural Variations (--fast or --no-sv is specified)')
bed_fpath = None
elif is_non_empty_file(bed_fpath):
logger.info(' Using existing BED-file: ' + bed_fpath)
elif not qconfig.forward_reads and not qconfig.interlaced_reads:
if not qconfig.reference_sam and not qconfig.reference_bam:
logger.info(' Will not search Structural Variations (needs paired-end reads)')
bed_fpath = None
qconfig.no_sv = True
if qconfig.create_icarus_html:
if is_non_empty_file(cov_fpath):
is_correct_file = check_cov_file(cov_fpath)
if is_correct_file:
logger.info(' Using existing reads coverage file: ' + cov_fpath)
if is_non_empty_file(physical_cov_fpath):
logger.info(' Using existing physical coverage file: ' + physical_cov_fpath)
else:
logger.info(' Will not calculate coverage (--fast or --no-html, or --no-icarus, or --space-efficient is specified)')
cov_fpath = None
physical_cov_fpath = None
if (is_non_empty_file(bed_fpath) or qconfig.no_sv) and \
(not qconfig.create_icarus_html or (is_non_empty_file(cov_fpath) and is_non_empty_file(physical_cov_fpath))):
required_files = []
n_jobs = min(qconfig.max_threads, len(contigs_fpaths) + 1)
max_threads_per_job = max(1, qconfig.max_threads // n_jobs)
sam_fpaths = qconfig.sam_fpaths or [None] * len(contigs_fpaths)
bam_fpaths = qconfig.bam_fpaths or [None] * len(contigs_fpaths)
parallel_align_args = [(contigs_fpath, output_dir, temp_output_dir, log_path, err_fpath, max_threads_per_job,
sam_fpaths[index], bam_fpaths[index], index) for index, contigs_fpath in enumerate(contigs_fpaths)]
if main_ref_fpath:
parallel_align_args.append((main_ref_fpath, output_dir, temp_output_dir, log_path, err_fpath,
max_threads_per_job, qconfig.reference_sam, qconfig.reference_bam, None, required_files, True))
correct_chr_names, sam_fpaths, bam_fpaths = run_parallel(align_single_file, parallel_align_args, n_jobs)
qconfig.sam_fpaths = sam_fpaths[:len(contigs_fpaths)]
qconfig.bam_fpaths = bam_fpaths[:len(contigs_fpaths)]
add_statistics_to_report(output_dir, contigs_fpaths, main_ref_fpath)
save_reads(output_dir)
if not main_ref_fpath:
return None, None, None
correct_chr_names = correct_chr_names[-1]
sam_fpath, bam_fpath = sam_fpaths[-1], bam_fpaths[-1]
qconfig.reference_sam = sam_fpath
qconfig.reference_bam = bam_fpath
if not required_files:
return bed_fpath, cov_fpath, physical_cov_fpath
if not all([sam_fpath, bam_fpath]):
logger.info(' Failed searching structural variations.')
return None, None, None
sam_sorted_fpath = get_safe_fpath(temp_output_dir, add_suffix(sam_fpath, 'sorted'))
bam_mapped_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_fpath, 'mapped'))
bam_sorted_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_mapped_fpath, 'sorted'))
if is_non_empty_file(sam_sorted_fpath):
logger.info(' Using existing sorted SAM-file: ' + sam_sorted_fpath)
else:
if not is_non_empty_file(bam_sorted_fpath):
sambamba_view(bam_fpath, bam_mapped_fpath, qconfig.max_threads, err_fpath, logger, filter_rule='not unmapped')
sort_bam(bam_mapped_fpath, bam_sorted_fpath, err_fpath, logger)
sambamba_view(bam_sorted_fpath, sam_sorted_fpath, qconfig.max_threads, err_fpath, logger)
if qconfig.create_icarus_html and (not is_non_empty_file(cov_fpath) or not is_non_empty_file(physical_cov_fpath)):
cov_fpath, physical_cov_fpath = get_coverage(temp_output_dir, main_ref_fpath, ref_name, bam_fpath, bam_sorted_fpath,
log_path, err_fpath, correct_chr_names, cov_fpath, physical_cov_fpath)
if not is_non_empty_file(bed_fpath) and not qconfig.no_sv:
if meta_ref_fpaths:
logger.info(' Splitting SAM-file by references...')
headers = []
seq_lengths = {}
with open(sam_fpath) as sam_file:
for line in sam_file:
if not line.startswith('@'):
break
if line.startswith('@SQ') and 'SN:' in line and 'LN:' in line:
seq_name = line.split('\tSN:')[1].split('\t')[0]
seq_length = int(line.split('\tLN:')[1].split('\t')[0])
seq_lengths[seq_name] = seq_length
headers.append(line.strip())
need_ref_splitting = False
ref_files = {}
if meta_ref_fpaths:
global ref_sam_fpaths
for cur_ref_fpath in meta_ref_fpaths:
cur_ref_name = qutils.name_from_fpath(cur_ref_fpath)
ref_sam_fpath = join(temp_output_dir, cur_ref_name + '.sam')
ref_sam_fpaths[cur_ref_fpath] = ref_sam_fpath
if is_non_empty_file(ref_sam_fpath):
logger.info(' Using existing split SAM-file for %s: %s' % (cur_ref_name, ref_sam_fpath))
ref_files[cur_ref_name] = None
else:
ref_sam_file = open(ref_sam_fpath, 'w')
if not headers[0].startswith('@SQ'):
ref_sam_file.write(headers[0] + '\n')
for h in (h for h in headers if h.startswith('@SQ') and 'SN:' in h):
seq_name = h.split('\tSN:')[1].split('\t')[0]
if seq_name in ref_labels and ref_labels[seq_name] == cur_ref_name:
ref_sam_file.write(h + '\n')
ref_sam_file.write(headers[-1] + '\n')
ref_files[cur_ref_name] = ref_sam_file
need_ref_splitting = True
trivial_deletions_fpath = \
search_trivial_deletions(temp_output_dir, sam_sorted_fpath, ref_files, ref_labels, seq_lengths, need_ref_splitting)
if get_gridss_fpath() and isfile(get_gridss_fpath()):
try:
gridss_sv_fpath = search_sv_with_gridss(main_ref_fpath, bam_mapped_fpath, meta_ref_fpaths, temp_output_dir, err_fpath)
qutils.cat_files([gridss_sv_fpath, trivial_deletions_fpath], bed_fpath)
except:
pass
if isfile(trivial_deletions_fpath) and not is_non_empty_file(bed_fpath):
shutil.copy(trivial_deletions_fpath, bed_fpath)
if not qconfig.no_sv:
if is_non_empty_file(bed_fpath):
logger.main_info(' Structural variations are in ' + bed_fpath)
else:
if isfile(bed_fpath):
logger.main_info(' No structural variations were found.')
else:
logger.main_info(' Failed searching structural variations.')
bed_fpath = None
if is_non_empty_file(cov_fpath):
logger.main_info(' Coverage distribution along the reference genome is in ' + cov_fpath)
else:
if not qconfig.create_icarus_html:
logger.main_info(' Failed to calculate coverage distribution')
cov_fpath = None
return bed_fpath, cov_fpath, physical_cov_fpath
def align_single_file(fpath, main_output_dir, output_dirpath, log_path, err_fpath, max_threads, sam_fpath=None, bam_fpath=None,
index=None, required_files=None, is_reference=False, alignment_only=False, using_reads='all'):
filename = qutils.name_from_fpath(fpath)
if not sam_fpath and bam_fpath:
sam_fpath = get_safe_fpath(output_dirpath, bam_fpath[:-4] + '.sam')
else:
sam_fpath = sam_fpath or join(output_dirpath, filename + '.sam')
bam_fpath = bam_fpath or get_safe_fpath(output_dirpath, sam_fpath[:-4] + '.bam')
if using_reads != 'all':
sam_fpath = join(output_dirpath, filename + '.' + using_reads + '.sam')
bam_fpath = sam_fpath.replace('.sam', '.bam')
if alignment_only or (is_reference and required_files and any(f.endswith('bed') for f in required_files)):
required_files.append(sam_fpath)
stats_fpath = get_safe_fpath(dirname(output_dirpath), filename + '.stat')
index_str = qutils.index_to_str(index) if index is not None else ''
reads_fpaths = qconfig.reads_fpaths
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
can_reuse = correct_chr_names is not None
if not can_reuse and not reads_fpaths:
return None, None, None
if correct_chr_names and (not required_files or all(isfile(fpath) for fpath in required_files)):
if not alignment_only:
if isfile(stats_fpath):
logger.info(' ' + index_str + 'Using existing flag statistics file ' + stats_fpath)
elif isfile(bam_fpath):
qutils.call_subprocess([sambamba_fpath('sambamba'), 'flagstat', '-t', str(max_threads), bam_fpath],
stdout=open(stats_fpath, 'w'), stderr=open(err_fpath, 'a'))
analyse_coverage(output_dirpath, fpath, correct_chr_names, bam_fpath, stats_fpath, err_fpath, logger)
calc_lap_score(reads_fpaths, sam_fpath, index, index_str, output_dirpath, fpath, filename, err_fpath)
if isfile(stats_fpath) or alignment_only:
return correct_chr_names, sam_fpath, bam_fpath
logger.info(' ' + index_str + 'Pre-processing reads...')
if is_non_empty_file(sam_fpath) and can_reuse:
logger.info(' ' + index_str + 'Using existing SAM-file: ' + sam_fpath)
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
elif is_non_empty_file(bam_fpath) and can_reuse:
logger.info(' ' + index_str + 'Using existing BAM-file: ' + bam_fpath)
sambamba_view(bam_fpath, sam_fpath, qconfig.max_threads, err_fpath, logger)
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
if (not correct_chr_names or not is_non_empty_file(sam_fpath)) and reads_fpaths:
if is_reference:
logger.info(' Running BWA for reference...')
else:
logger.info(' ' + index_str + 'Running BWA...')
# use absolute paths because we will change workdir
fpath = abspath(fpath)
sam_fpath = abspath(sam_fpath)
prev_dir = os.getcwd()
os.chdir(output_dirpath)
bwa_index(fpath, err_fpath, logger)
sam_fpaths = align_reads(fpath, sam_fpath, using_reads, main_output_dir, err_fpath, max_threads)
if len(sam_fpaths) > 1:
merge_sam_files(sam_fpaths, sam_fpath, bam_fpath, main_output_dir, max_threads, err_fpath)
elif len(sam_fpaths) == 1:
shutil.move(sam_fpaths[0], sam_fpath)
sambamba_view(sam_fpath, bam_fpath, max_threads, err_fpath, logger, filter_rule=None)
logger.info(' ' + index_str + 'Done.')
os.chdir(prev_dir)
if not is_non_empty_file(sam_fpath):
logger.error(' Failed running BWA for ' + fpath + '. See ' + log_path + ' for information.')
return None, None, None
correct_chr_names = get_correct_names_for_chroms(output_dirpath, fpath, sam_fpath, err_fpath, reads_fpaths, logger, is_reference)
elif not correct_chr_names or not is_non_empty_file(sam_fpath):
return None, None, None
if is_reference:
logger.info(' Sorting SAM-file for reference...')
else:
logger.info(' ' + index_str + 'Sorting SAM-file...')
if can_reuse and is_non_empty_file(bam_fpath) and all_read_names_correct(sam_fpath):
logger.info(' ' + index_str + 'Using existing BAM-file: ' + bam_fpath)
else:
correct_sam_fpath = join(output_dirpath, filename + '.correct.sam') # write in output dir
sam_fpath = clean_read_names(sam_fpath, correct_sam_fpath)
sambamba_view(correct_sam_fpath, bam_fpath, max_threads, err_fpath, logger, filter_rule=None)
qutils.assert_file_exists(bam_fpath, 'bam file')
if not alignment_only:
if isfile(stats_fpath):
logger.info(' ' + index_str + 'Using existing flag statistics file ' + stats_fpath)
elif isfile(bam_fpath):
qutils.call_subprocess([sambamba_fpath('sambamba'), 'flagstat', '-t', str(max_threads), bam_fpath],
stdout=open(stats_fpath, 'w'), stderr=open(err_fpath, 'a'))
analyse_coverage(output_dirpath, fpath, correct_chr_names, bam_fpath, stats_fpath, err_fpath, logger)
calc_lap_score(reads_fpaths, sam_fpath, index, index_str, output_dirpath, fpath, filename, err_fpath)
if is_reference:
logger.info(' Analysis for reference is finished.')
else:
logger.info(' ' + index_str + 'Analysis is finished.')
return correct_chr_names, sam_fpath, bam_fpath
def run_processing_reads(contigs_fpaths, main_ref_fpath, meta_ref_fpaths, ref_labels, temp_output_dir, output_dir,
log_path, err_fpath):
required_files = []
bed_fpath, cov_fpath, physical_cov_fpath = None, None, None
if main_ref_fpath:
ref_name = qutils.name_from_fpath(main_ref_fpath)
bed_fpath = qconfig.bed or join(output_dir, ref_name + '.bed')
cov_fpath = qconfig.cov_fpath or join(output_dir, ref_name + '.cov')
physical_cov_fpath = qconfig.phys_cov_fpath or join(output_dir, ref_name + '.physical.cov')
required_files = [bed_fpath, cov_fpath, physical_cov_fpath]
if qconfig.no_sv:
logger.info(' Will not search Structural Variations (--fast or --no-sv is specified)')
bed_fpath = None
elif is_non_empty_file(bed_fpath):
logger.info(' Using existing BED-file: ' + bed_fpath)
elif not qconfig.forward_reads and not qconfig.interlaced_reads:
if not qconfig.reference_sam and not qconfig.reference_bam:
logger.info(' Will not search Structural Variations (needs paired-end reads)')
bed_fpath = None
qconfig.no_sv = True
if qconfig.create_icarus_html:
if is_non_empty_file(cov_fpath):
is_correct_file = check_cov_file(cov_fpath)
if is_correct_file:
logger.info(' Using existing reads coverage file: ' + cov_fpath)
if is_non_empty_file(physical_cov_fpath):
logger.info(' Using existing physical coverage file: ' + physical_cov_fpath)
else:
logger.info(' Will not calculate coverage (--fast or --no-html, or --no-icarus, or --space-efficient is specified)')
cov_fpath = None
physical_cov_fpath = None
if (is_non_empty_file(bed_fpath) or qconfig.no_sv) and \
(not qconfig.create_icarus_html or (is_non_empty_file(cov_fpath) and is_non_empty_file(physical_cov_fpath))):
required_files = []
n_jobs = min(qconfig.max_threads, len(contigs_fpaths) + 1)
max_threads_per_job = max(1, qconfig.max_threads // n_jobs)
sam_fpaths = qconfig.sam_fpaths or [None] * len(contigs_fpaths)
bam_fpaths = qconfig.bam_fpaths or [None] * len(contigs_fpaths)
parallel_align_args = [(contigs_fpath, output_dir, temp_output_dir, log_path, err_fpath, max_threads_per_job,
sam_fpaths[index], bam_fpaths[index], index) for index, contigs_fpath in enumerate(contigs_fpaths)]
if main_ref_fpath:
parallel_align_args.append((main_ref_fpath, output_dir, temp_output_dir, log_path, err_fpath,
max_threads_per_job, qconfig.reference_sam, qconfig.reference_bam, None, required_files, True))
correct_chr_names, sam_fpaths, bam_fpaths = run_parallel(align_single_file, parallel_align_args, n_jobs)
qconfig.sam_fpaths = sam_fpaths[:len(contigs_fpaths)]
qconfig.bam_fpaths = bam_fpaths[:len(contigs_fpaths)]
add_statistics_to_report(output_dir, contigs_fpaths, main_ref_fpath)
save_reads(output_dir)
if not main_ref_fpath:
return None, None, None
correct_chr_names = correct_chr_names[-1]
sam_fpath, bam_fpath = sam_fpaths[-1], bam_fpaths[-1]
qconfig.reference_sam = sam_fpath
qconfig.reference_bam = bam_fpath
if not required_files:
return bed_fpath, cov_fpath, physical_cov_fpath
if not all([sam_fpath, bam_fpath]):
logger.info(' Failed searching structural variations.')
return None, None, None
sam_sorted_fpath = get_safe_fpath(temp_output_dir, add_suffix(sam_fpath, 'sorted'))
bam_mapped_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_fpath, 'mapped'))
bam_sorted_fpath = get_safe_fpath(temp_output_dir, add_suffix(bam_mapped_fpath, 'sorted'))
if is_non_empty_file(sam_sorted_fpath):
logger.info(' Using existing sorted SAM-file: ' + sam_sorted_fpath)
else:
if not is_non_empty_file(bam_sorted_fpath):
sambamba_view(bam_fpath, bam_mapped_fpath, qconfig.max_threads, err_fpath, logger, filter_rule='not unmapped')
sort_bam(bam_mapped_fpath, bam_sorted_fpath, err_fpath, logger)
sambamba_view(bam_sorted_fpath, sam_sorted_fpath, qconfig.max_threads, err_fpath, logger)
if qconfig.create_icarus_html and (not is_non_empty_file(cov_fpath) or not is_non_empty_file(physical_cov_fpath)):
cov_fpath, physical_cov_fpath = get_coverage(temp_output_dir, main_ref_fpath, ref_name, bam_fpath, bam_sorted_fpath,
log_path, err_fpath, correct_chr_names, cov_fpath, physical_cov_fpath)
if not is_non_empty_file(bed_fpath) and not qconfig.no_sv:
if meta_ref_fpaths:
logger.info(' Splitting SAM-file by references...')
headers = []
seq_lengths = {}
with open(sam_fpath) as sam_file:
for line in sam_file:
if not line.startswith('@'):
break
if line.startswith('@SQ') and 'SN:' in line and 'LN:' in line:
seq_name = line.split('\tSN:')[1].split('\t')[0]
seq_length = int(line.split('\tLN:')[1].split('\t')[0])
seq_lengths[seq_name] = seq_length
headers.append(line.strip())
need_ref_splitting = False
ref_files = {}
if meta_ref_fpaths:
global ref_sam_fpaths
for cur_ref_fpath in meta_ref_fpaths:
cur_ref_name = qutils.name_from_fpath(cur_ref_fpath)
ref_sam_fpath = join(temp_output_dir, cur_ref_name + '.sam')
ref_sam_fpaths[cur_ref_fpath] = ref_sam_fpath
if is_non_empty_file(ref_sam_fpath):
logger.info(' Using existing split SAM-file for %s: %s' % (cur_ref_name, ref_sam_fpath))
ref_files[cur_ref_name] = None
else:
ref_sam_file = open(ref_sam_fpath, 'w')
if not headers[0].startswith('@SQ'):
ref_sam_file.write(headers[0] + '\n')
for h in (h for h in headers if h.startswith('@SQ') and 'SN:' in h):
seq_name = h.split('\tSN:')[1].split('\t')[0]
if seq_name in ref_labels and ref_labels[seq_name] == cur_ref_name:
ref_sam_file.write(h + '\n')
ref_sam_file.write(headers[-1] + '\n')
ref_files[cur_ref_name] = ref_sam_file
need_ref_splitting = True
trivial_deletions_fpath = \
search_trivial_deletions(temp_output_dir, sam_sorted_fpath, ref_files, ref_labels, seq_lengths, need_ref_splitting)
if get_gridss_fpath() and isfile(get_gridss_fpath()):
try:
gridss_sv_fpath = search_sv_with_gridss(main_ref_fpath, bam_mapped_fpath, meta_ref_fpaths, temp_output_dir, err_fpath)
qutils.cat_files([gridss_sv_fpath, trivial_deletions_fpath], bed_fpath)
except:
pass
if isfile(trivial_deletions_fpath) and not is_non_empty_file(bed_fpath):
shutil.copy(trivial_deletions_fpath, bed_fpath)
if not qconfig.no_sv:
if is_non_empty_file(bed_fpath):
logger.main_info(' Structural variations are in ' + bed_fpath)
else:
if isfile(bed_fpath):
logger.main_info(' No structural variations were found.')
else:
logger.main_info(' Failed searching structural variations.')
bed_fpath = None
if is_non_empty_file(cov_fpath):
logger.main_info(' Coverage distribution along the reference genome is in ' + cov_fpath)
else:
if not qconfig.create_icarus_html:
logger.main_info(' Failed to calculate coverage distribution')
cov_fpath = None
return bed_fpath, cov_fpath, physical_cov_fpath