def main():
# read params
args = parse_arguments()
log.info('Initializing and making output directory...')
mkdir_p(args.out_dir)
log.info('Do naive overlap...')
overlap_peak = naive_overlap(
args.prefix,
args.peak1,
args.peak2,
args.peak_pooled,
args.peak_type,
args.nonamecheck,
args.mem_gb,
args.out_dir,
)
log.info('Blacklist-filtering peaks...')
bfilt_overlap_peak = blacklist_filter(overlap_peak, args.blacklist,
args.regex_bfilt_peak_chr_name,
args.out_dir)
log.info('Checking if output is empty...')
assert_file_not_empty(bfilt_overlap_peak)
log.info('Converting peak to bigbed...')
peak_to_bigbed(bfilt_overlap_peak, args.peak_type, args.chrsz, args.mem_gb,
args.out_dir)
log.info('Converting peak to starch...')
peak_to_starch(bfilt_overlap_peak, args.out_dir)
log.info('Converting peak to hammock...')
peak_to_hammock(bfilt_overlap_peak, args.mem_gb, args.out_dir)
if args.ta: # if TAG-ALIGN is given
if args.fraglen: # chip-seq
log.info('Shifted FRiP with fragment length...')
frip_shifted(args.ta, bfilt_overlap_peak, args.chrsz, args.fraglen,
args.out_dir)
else: # atac-seq
log.info('FRiP without fragment length...')
frip(args.ta, bfilt_overlap_peak, args.out_dir)
log.info('List all files in output directory...')
ls_l(args.out_dir)
log.info('All done.')
def main():
# read params
args = parse_arguments()
log.info('Initializing and making output directory...')
mkdir_p(args.out_dir)
log.info('Do IDR...')
idr_peak, idr_plot, idr_out_gz, idr_stdout = idr(
args.prefix,
args.peak1, args.peak2, args.peak_pooled, args.peak_type,
args.chrsz,
args.idr_thresh, args.idr_rank, args.mem_gb, args.out_dir,
)
log.info('Checking if output is empty...')
assert_file_not_empty(idr_peak, help=
'No IDR peaks found. IDR threshold might be too stringent '
'or replicates have very poor concordance.')
log.info('Blacklist-filtering peaks...')
bfilt_idr_peak = blacklist_filter(
idr_peak, args.blacklist, args.regex_bfilt_peak_chr_name, args.out_dir)
log.info('Converting peak to bigbed...')
peak_to_bigbed(bfilt_idr_peak, args.peak_type, args.chrsz,
args.mem_gb, args.out_dir)
log.info('Converting peak to starch...')
peak_to_starch(bfilt_idr_peak, args.out_dir)
log.info('Converting peak to hammock...')
peak_to_hammock(bfilt_idr_peak, args.mem_gb, args.out_dir)
if args.ta: # if TAG-ALIGN is given
if args.fraglen: # chip-seq
log.info('Shifted FRiP with fragment length...')
frip_shifted(args.ta, bfilt_idr_peak,
args.chrsz, args.fraglen, args.out_dir)
else: # atac-seq
log.info('FRiP without fragment length...')
frip(args.ta, bfilt_idr_peak, args.out_dir)
log.info('List all files in output directory...')
ls_l(args.out_dir)
log.info('All done.')
def main():
# read params
args = parse_arguments()
log.info('Initializing and making output directory...')
mkdir_p(args.out_dir)
log.info('Blacklist-filtering peaks...')
bfilt_peak = blacklist_filter(args.peak, args.blacklist,
args.regex_bfilt_peak_chr_name, args.out_dir)
log.info('Checking if output is empty...')
assert_file_not_empty(bfilt_peak)
log.info('Converting peak to bigbed...')
peak_to_bigbed(bfilt_peak, args.peak_type, args.chrsz, args.mem_gb,
args.out_dir)
log.info('Converting peak to starch...')
peak_to_starch(bfilt_peak, args.out_dir)
log.info('Converting peak to hammock...')
peak_to_hammock(bfilt_peak, args.mem_gb, args.out_dir)
log.info('Shifted FRiP with fragment length...')
frip_qc = frip_shifted(args.ta, bfilt_peak, args.chrsz, args.fraglen,
args.out_dir)
log.info('Calculating (blacklist-filtered) peak region size QC/plot...')
region_size_qc, region_size_plot = get_region_size_metrics(bfilt_peak)
log.info('Calculating number of peaks (blacklist-filtered)...')
num_peak_qc = get_num_peaks(bfilt_peak)
log.info('List all files in output directory...')
ls_l(args.out_dir)
log.info('All done.')
def test_starch_and_unstarch(tmp_path):
"""Cannot use md5hash of output starch file since it includes timestamp.
So unstarch it and calculate md5 hash.
This is actually an integration test of starch/unstarch.
Two functions are tested together:
- peak_to_starch(): cannot control starch's timestamp inclusion.
- starch_to_bed_gz(): this function gzips with -n (excluding timestamp).
"""
peak = tmp_path / 'idr_peak.gz'
peak.write_text(IDR_PEAK_FILE_CONTENTS)
starch = peak_to_starch(peak, tmp_path)
bed_gz = starch_to_bed_gz(starch, tmp_path)
with open(bed_gz, 'rb') as fp:
assert hashlib.md5(fp.read()).hexdigest() == MD5_HASH_IDR_PEAK_UNSTARCHED
def main():
# read params
args = parse_arguments()
log.info('Initializing and making output directory...')
mkdir_p(args.out_dir)
log.info('Reproducibility QC...')
# description for variables
# N: list of number of peaks in peak files from pseudo replicates
# Nt: top number of peaks in peak files
# from true replicates (rep-x_vs_rep-y)
# Np: number of peaks in peak files from pooled pseudo replicate
N = [get_num_lines(peak) for peak in args.peaks_pr]
if len(args.peaks):
# multiple replicate case
num_rep = infer_n_from_nC2(len(args.peaks))
num_peaks_tr = [get_num_lines(peak) for peak in args.peaks]
Nt = max(num_peaks_tr)
Np = get_num_lines(args.peak_ppr)
rescue_ratio = float(max(Np, Nt)) / float(min(Np, Nt))
self_consistency_ratio = float(max(N)) / float(min(N))
Nt_idx = num_peaks_tr.index(Nt)
label_tr = infer_pair_label_from_idx(num_rep, Nt_idx)
conservative_set = label_tr
conservative_peak = args.peaks[Nt_idx]
N_conservative = Nt
if Nt > Np:
optimal_set = conservative_set
optimal_peak = conservative_peak
N_optimal = N_conservative
else:
optimal_set = "pooled-pr1_vs_pooled-pr2"
optimal_peak = args.peak_ppr
N_optimal = Np
else:
# single replicate case
num_rep = 1
Nt = 0
Np = 0
rescue_ratio = 0.0
self_consistency_ratio = 1.0
conservative_set = 'rep1-pr1_vs_rep1-pr2'
conservative_peak = args.peaks_pr[0]
N_conservative = N[0]
optimal_set = conservative_set
optimal_peak = conservative_peak
N_optimal = N_conservative
reproducibility = 'pass'
if rescue_ratio > 2.0 or self_consistency_ratio > 2.0:
reproducibility = 'borderline'
if rescue_ratio > 2.0 and self_consistency_ratio > 2.0:
reproducibility = 'fail'
log.info('Writing optimal/conservative peak files...')
optimal_peak_file = os.path.join(
args.out_dir, '{}optimal_peak.{}.gz'.format(
(args.prefix + '.') if args.prefix else '', args.peak_type))
conservative_peak_file = os.path.join(
args.out_dir, '{}conservative_peak.{}.gz'.format(
(args.prefix + '.') if args.prefix else '', args.peak_type))
copy_f_to_f(optimal_peak, optimal_peak_file)
copy_f_to_f(conservative_peak, conservative_peak_file)
if args.chrsz:
log.info('Converting peak to bigbed...')
peak_to_bigbed(optimal_peak_file, args.peak_type, args.chrsz,
args.out_dir)
peak_to_bigbed(conservative_peak_file, args.peak_type, args.chrsz,
args.out_dir)
log.info('Converting peak to starch...')
peak_to_starch(optimal_peak_file, args.out_dir)
peak_to_starch(conservative_peak_file, args.out_dir)
log.info('Converting peak to hammock...')
peak_to_hammock(optimal_peak_file, args.out_dir)
peak_to_hammock(conservative_peak_file, args.out_dir)
log.info('Writing reproducibility QC log...')
if args.prefix:
reproducibility_qc = '{}.reproducibility.qc'.format(args.prefix)
else:
reproducibility_qc = 'reproducibility.qc'
reproducibility_qc = os.path.join(args.out_dir, reproducibility_qc)
with open(reproducibility_qc, 'w') as fp:
header = '{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\n'.format(
'Nt',
'\t'.join(['N{}'.format(i + 1) for i in range(num_rep)]),
'Np',
'N_opt',
'N_consv',
'opt_set',
'consv_set',
'rescue_ratio',
'self_consistency_ratio',
'reproducibility',
)
line = '{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\n'.format(
Nt, '\t'.join([str(i) for i in N]), Np, N_optimal, N_conservative,
optimal_set, conservative_set, rescue_ratio,
self_consistency_ratio, reproducibility)
fp.write(header)
fp.write(line)
log.info('Calculating (optimal) peak region size QC/plot...')
region_size_qc, region_size_plot = get_region_size_metrics(
optimal_peak_file)
log.info('Calculating number of peaks (optimal)...')
get_num_peaks(optimal_peak_file)
log.info('All done.')