def main():
"""
Generage homog files. Given a blocks file and pat[s],
count the number of U,X,M reads for each block for each file
"""
args = parse_args()
if args.nr_bits not in (8 , 16):
raise IllegalArgumentError('nr_bits must be in {8, 16}')
if args.rlen < 3:
raise IllegalArgumentError('rlen must be >= 3')
if args.thresholds is not None:
th = args.thresholds.split(',')
if not len(th) == 2: # and th[0].is_number():
raise IllegalArgumentError('Invalid thresholds')
th = float(th[0]), float(th[1])
if not (1 > th[1] > th[0] > 0):
raise IllegalArgumentError('Invalid thresholds')
# make sure homog tool is valid:
validate_local_exe(homog_tool)
pats = args.input_files
validate_file_list(pats, '.pat.gz')
outdir, prefix = parse_outdir_prefix(args)
# load blocks:
blocks_df = load_blocks_file(args.blocks_file)
is_nice, msg = is_block_file_nice(blocks_df)
if not is_nice:
homog_log(msg)
raise IllegalArgumentError(f'Invalid blocks file: {args.blocks_file}')
for pat in sorted(pats):
homog_process(pat, blocks_df, args, outdir, prefix)
def view_gr(pat, args, get_cmd=False):
validate_single_file(pat, '.pat.gz')
gr = GenomicRegion(args)
if gr.is_whole():
s = 1
e = gr.genome.get_nr_sites() + 1
cmd = f'gunzip -c {pat} '
else:
s, e = gr.sites
ms = max(1, s - MAX_PAT_LEN)
cmd = f'tabix {pat} {gr.chrom}:{ms}-{e - 1} '
view_flags = set_view_flags(args)
cmd += f' | {cview_tool} --sites "{s}\t{e}" ' + view_flags
if hasattr(
args,
'sub_sample') and args.sub_sample is not None: # sub-sample reads
validate_local_exe(pat_sampler)
cmd += f' | {pat_sampler} {args.sub_sample} '
if not gr.is_whole():
cmd += f' | sort -k2,2n -k3,3 '
cmd += f' | {collapse_pat_script} - '
if get_cmd:
return cmd
if args.out_path is not None:
cmd += f' > {args.out_path}'
subprocess_wrap_sigpipe(cmd)
def add_bed_to_cpgs(site_file, genome, out_path=None):
validate_local_exe(add_loci_tool)
g = GenomeRefPaths(genome)
cmd = f'cat {site_file} | {add_loci_tool} {g.dict_path} {g.chrom_cpg_sizes}'
if (out_path is not None) and out_path != sys.stdout:
cmd += f' > {out_path}'
subprocess.check_call(cmd, shell=True)
def main():
"""
Generate a beta file from a pat file
"""
args = parse_args()
validate_local_exe(pat2beta_tool)
for pat in args.pat_paths:
pat2beta(pat, args.out_dir, args, args.force)
def main():
"""
Segment the genome, or a subset region, to homogenously methylated blocks.
Input: one or more beta files to segment
Output: blocks file (BED format + startCpG, endCpG columns)
"""
args = parse_args()
validate_local_exe(segment_tool)
betas = parse_betas_input(args)
SegmentByChunks(args, betas).run()
def main():
""" view pat file with the c++ engine """
parser = parse_args()
args = parser.parse_args()
# validate input file
pat = args.pat
validate_single_file(pat)
if (args.sub_sample is not None) and (args.sub_sample < 0):
parser.error('[wt view] sub-sampling rate must be >= 0')
validate_local_exe(cview_tool)
cview(pat, args)
def main():
"""
Add to bam file an extra field, YI:Z:{nr_meth},{nr_unmeth},
to count Cytosine retention at CpG context.
"""
parser = argparse.ArgumentParser(description=main.__doc__)
parser = add_args(parser)
parser = add_cpg_args(parser)
args = parser.parse_args()
validate_local_exe(add_cpg_count_tool)
for bam in args.bam:
BamMethylData(args, bam).start_threads()
def main():
"""
Run the WGBS pipeline to generate pat & beta files out of an input bam file
"""
parser = add_args_snp_splitt()
args = parse_args_snp_split(parser)
# validate output dir:
if not op.isdir(args.out_dir):
raise IllegalArgumentError(f'Invalid output dir: {args.out_dir}')
validate_local_exe(allele_split_tool)
for bam in [args.bam]:
if not validate_bam(bam):
eprint(f'[wt bam2pat] Skipping {bam}')
continue
pat = op.join(args.out_dir, op.basename(bam)[:-4] + PAT_SUFF)
if not delete_or_skip(pat, args.force):
continue
SNPSplit(args, bam)
def main():
"""
Run the WGBS pipeline to generate pat & beta files out of an input bam file
"""
parser = argparse.ArgumentParser(description=main.__doc__)
parser = add_args(parser)
args = parse_args(parser)
# validate output dir:
if not op.isdir(args.out_dir):
raise IllegalArgumentError(f'Invalid output dir: {args.out_dir}')
validate_local_exe(match_maker_tool)
validate_local_exe(patter_tool)
for bam in args.bam:
if not validate_bam(bam):
eprint(f'[wt bam2pat] Skipping {bam}')
continue
pat = op.join(args.out_dir, op.basename(bam)[:-4] + PAT_SUFF)
if not delete_or_skip(pat, args.force):
continue
Bam2Pat(args, bam)
def view_bed(pat, args):
# assume columns 4-5 of args.bed_file are startCpG, endCpG:
bpath = args.bed_file
# validate blocks file. If it's long, and starts with "chr1", use gunzip instead of tabix.
df = load_blocks_file(bpath, nrows=1e6)
if df.shape[0] == 1e6 and df.iloc[0, 0] in ('1', 'chr1'):
tabix_cmd = f'gunzip -c {pat} '
else:
# extended blocks:
tabix_cmd = 'gunzip -c' if bpath.endswith('.gz') else 'cat'
tabix_cmd += f' {bpath} | {cview_extend_blocks_script} | tabix -R - {pat} '
view_flags = set_view_flags(args)
cmd = tabix_cmd + f' | {cview_tool} {view_flags} --blocks_path {bpath}'
if args.sub_sample is not None: # sub-sample reads
validate_local_exe(pat_sampler)
cmd += f' | {pat_sampler} {args.sub_sample} '
cmd += f' | sort -k2,2n -k3,3 | {collapse_pat_script} - '
if args.out_path is not None:
cmd += f' > {args.out_path}'
subprocess_wrap_sigpipe(cmd)