def input_parser(filepath):
bamfiles_1 = get_data_block(filepath, "rep1")
bamfiles_1 = map(npath, bamfiles_1)
bamfiles_2 = get_data_block(filepath, "rep2")
bamfiles_2 = map(npath, bamfiles_2)
# genome is optional, so if we get an empty list
# we set it to None, otherwise we normalise the path
genome = get_data_block(filepath, "genome")
genome = npath(genome) if genome else None
# the chrom sizes are not optional, but right now it's undefined
# what happens if the user doesn't specify them, or specifies more
# than one. So we just relay whatever we got from the file.
chrom_sizes = npath(get_data_block(filepath, "chrom_sizes"))
chrom_sizes = npath(chrom_sizes) if chrom_sizes else chrom_sizes
inputs1 = get_data_block(filepath, "inputs1")
inputs1 = map(npath, inputs1)
inputs2 = get_data_block(filepath, "inputs2")
inputs2 = map(npath, inputs2)
dims = [len(bamfiles_1), len(bamfiles_2)]
if not inputs1 and not inputs2:
inputs = None
else:
inputs = inputs1 + inputs2
return bamfiles_1 + bamfiles_2, genome, chrom_sizes, inputs, dims
def read_enrichment(self, enrichment_files, threshold=1):
"""
Reads current output of motif enrichment analysis to get gene targets.
*Keyword arguments:*
- enrichment_files -- One string, or a list of strings, representing enrichment file paths.
- threshold -- P-value threshold for motif acceptance.
"""
if isinstance(enrichment_files, list):
file_list = [
filename for pattern in enrichment_files
for filename in glob.glob(npath(pattern))
]
else:
file_list = glob.glob(npath(enrichment_files))
# reading networks
for filename in file_list:
# use last dir name as name for condition
condition = os.path.dirname(filename)
condition = condition.split("/")[-1]
self.conditions.append(condition)
network = {}
f = open(filename, "r")
# skip header
next(f)
for line in f:
line = line.strip("\n")
values = line.split("\t")
motif = values[0]
if motif in self.motifs_map:
p_value = float(values[2])
genes = values[9].split(",")
if threshold >= p_value:
network[motif] = genes
if motif in self.motifs_enrichment:
self.motifs_enrichment[motif][condition] = p_value
else:
self.motifs_enrichment[motif] = {condition: p_value}
else:
print("motif not found: " + motif)
self.networks[condition] = network
f.close()
def write_enrichment(self, out_file, threshold=1):
"""
Writes enrichment table for network generation.
*Keyword arguments:*
- out_file -- Output file name.
- threshold -- P-value threshold for motif acceptance.
"""
f = open(npath(out_file), "w")
f.write("\t" + ("\t".join(self.conditions)) + "\n")
for v in self.motifs_enrichment:
values = self.motifs_enrichment[v]
filter_p = False
p_values = []
for c in self.conditions:
if c in values:
pvalue = values[c]
p_values.append(str(pvalue))
if pvalue <= threshold:
filter_p = True
else:
p_values.append("1")
if filter_p and (
v in self.motifs_map) and self.motifs_map[v].gene_names:
genes = "|".join(self.motifs_map[v].gene_names)
f.write(v + "|" + genes + "\t" + ("\t".join(p_values)) + "\n")
def merge_output(bamfiles, dims, options, no_bw_files, chrom_sizes):
for i in range(len(bamfiles)):
rep = i if i < dims[0] else i - dims[0]
sig = 1 if i < dims[0] else 2
temp_bed = npath(options.name + '-s%s-rep%s_temp.bed' % (sig, rep))
files = [options.name + '-' + str(j) + '-s%s-rep%s.bw' %(sig, rep) for j in no_bw_files]
if len(no_bw_files) > len(bamfiles):
files = filter(lambda x: isfile(x), files)
t = ['bigWigMerge'] + files + [temp_bed]
c = " ".join(t)
os.system(c)
os.system("LC_COLLATE=C sort -k1,1 -k2,2n " + temp_bed + ' > ' + temp_bed +'.sort')
t = ['bedGraphToBigWig', temp_bed + '.sort', chrom_sizes, options.name + '-s%s-rep%s.bw' % (sig, rep)]
c = " ".join(t)
os.system(c)
for f in files:
os.remove(f)
os.remove(temp_bed)
os.remove(temp_bed + ".sort")
else:
ftarget = [options.name + '-s%s-rep%s.bw' %(sig, rep) for j in no_bw_files]
for i in range(len(ftarget)):
c = ['mv', files[i], ftarget[i]]
c = " ".join(c)
os.system(c)
def write_enrichment(self, out_file, threshold=1):
"""
Writes enrichment table for network generation.
*Keyword arguments:*
- out_file -- Output file name.
- threshold -- P-value threshold for motif acceptance.
"""
f = open(npath(out_file), "w")
f.write("\t" + ("\t".join(self.conditions)) + "\n")
for v in self.motifs_enrichment:
values = self.motifs_enrichment[v]
filter_p = False
p_values = []
for c in self.conditions:
if c in values:
pvalue = values[c]
p_values.append(str(pvalue))
if pvalue <= threshold:
filter_p = True
else:
p_values.append("1")
if filter_p and (v in self.motifs_map) and self.motifs_map[v].gene_names:
genes = "|".join(self.motifs_map[v].gene_names)
f.write(v + "|" + genes + "\t" + ("\t".join(p_values)) + "\n")
def read_mtf(self, mtf_filenames):
"""
Reads TF annotation in mtf (internal format; check manual) format.
*Keyword arguments:*
- mtf_filenames -- A string, or a list of strings, representing .mtf file paths.
"""
if not isinstance(mtf_filenames, list):
mtf_filenames = [mtf_filenames]
file_list = [
filename for pattern in mtf_filenames
for filename in glob.glob(npath(pattern))
]
# Iterating over the file name list
for filename in file_list:
database = os.path.splitext(os.path.basename(filename))[0]
# Opening MTF file
mtf_file = open(filename, "r")
# Reading file
for line in mtf_file:
# Processing line
line_list = line.strip().split("\t")
tf_id = line_list[0].strip()
name = line_list[1].strip()
version = line_list[2].strip()
gene_names = line_list[3].strip().split("+")
tf_class = line_list[4].strip()
uniprot_ids = line_list[5].strip().split(";")
data_source = line_list[6].strip()
tax_group = line_list[7].strip()
species = line_list[8].strip()
threshold_list = line_list[9].strip().split(",")
fpr_list = [0.005, 0.001, 0.0005, 0.0001, 0.00005, 0.00001]
thresholds = {}
for i in range(0, 6):
thresholds[fpr_list[i]] = float(threshold_list[i])
self.add(
MotifAnnotation(tf_id, name, database, version, gene_names,
tf_class, uniprot_ids, data_source,
tax_group, species, thresholds))
# Termination
mtf_file.close()
def read_enrichment(self, enrichment_files, threshold=1):
"""
Reads current output of motif enrichment analysis to get gene targets.
*Keyword arguments:*
- enrichment_files -- One string, or a list of strings, representing enrichment file paths.
- threshold -- P-value threshold for motif acceptance.
"""
if isinstance(enrichment_files, list):
file_list = [filename for pattern in enrichment_files for filename in glob.glob(npath(pattern))]
else:
file_list = glob.glob(npath(enrichment_files))
# reading networks
for filename in file_list:
# use last dir name as name for condition
condition = os.path.dirname(filename)
condition = condition.split("/")[-1]
self.conditions.append(condition)
network = {}
f = open(filename, "r")
# skip header
next(f)
for line in f:
line = line.strip("\n")
values = line.split("\t")
motif = values[0]
if motif in self.motifs_map:
p_value = float(values[2])
genes = values[9].split(",")
if threshold >= p_value:
network[motif] = genes
if motif in self.motifs_enrichment:
self.motifs_enrichment[motif][condition] = p_value
else:
self.motifs_enrichment[motif] = {condition: p_value}
else:
print("motif not found: " + motif)
self.networks[condition] = network
f.close()
def read_mtf(self, mtf_filenames):
"""
Reads TF annotation in mtf (internal format; check manual) format.
*Keyword arguments:*
- mtf_filenames -- A string, or a list of strings, representing .mtf file paths.
"""
if not isinstance(mtf_filenames, list):
mtf_filenames = [mtf_filenames]
file_list = [filename for pattern in mtf_filenames for filename in glob.glob(npath(pattern))]
# Iterating over the file name list
for filename in file_list:
database = os.path.splitext(os.path.basename(filename))[0]
# Opening MTF file
mtf_file = open(filename, "r")
# Reading file
for line in mtf_file:
# Processing line
line_list = line.strip().split("\t")
tf_id = line_list[0].strip()
name = line_list[1].strip()
version = line_list[2].strip()
gene_names = line_list[3].strip().split("+")
tf_class = line_list[4].strip()
uniprot_ids = line_list[5].strip().split(";")
data_source = line_list[6].strip()
tax_group = line_list[7].strip()
species = line_list[8].strip()
threshold_list = line_list[9].strip().split(",")
fpr_list = [0.005, 0.001, 0.0005, 0.0001, 0.00005, 0.00001]
thresholds = {}
for i in range(0, 6):
thresholds[fpr_list[i]] = float(threshold_list[i])
self.add(MotifAnnotation(tf_id, name, database, version, gene_names, tf_class, uniprot_ids, data_source,
tax_group, species, thresholds))
# Termination
mtf_file.close()
def merge_output(bamfiles, dims, options, no_bw_files, chrom_sizes):
for i in range(len(bamfiles)):
rep = i if i < dims[0] else i - dims[0]
sig = 1 if i < dims[0] else 2
temp_bed = npath(options.name + '-s%s-rep%s_temp.bed' % (sig, rep))
files = [
options.name + '-' + str(j) + '-s%s-rep%s.bw' % (sig, rep)
for j in no_bw_files
]
if len(no_bw_files) > len(bamfiles):
files = filter(lambda x: isfile(x), files)
t = ['bigWigMerge'] + files + [temp_bed]
c = " ".join(t)
os.system(c)
os.system("LC_COLLATE=C sort -k1,1 -k2,2n " + temp_bed + ' > ' +
temp_bed + '.sort')
t = [
'bedGraphToBigWig', temp_bed + '.sort', chrom_sizes,
options.name + '-s%s-rep%s.bw' % (sig, rep)
]
c = " ".join(t)
os.system(c)
for f in files:
os.remove(f)
os.remove(temp_bed)
os.remove(temp_bed + ".sort")
else:
ftarget = [
options.name + '-s%s-rep%s.bw' % (sig, rep)
for j in no_bw_files
]
for i in range(len(ftarget)):
c = ['mv', files[i], ftarget[i]]
c = " ".join(c)
os.system(c)
def read_mtf(self, mtf_filenames):
"""
Reads TF annotation in mtf (internal format; check manual) format.
*Keyword arguments:*
- file_name_list -- A string, or a list of strings, representing .mtf file paths.
"""
if isinstance(mtf_filenames, list):
file_list = [filename for pattern in mtf_filenames for filename in glob.glob(npath(pattern))]
else:
file_list = glob.glob(npath(mtf_filenames))
# Iterating over the file name list
for filename in file_list:
# Opening MTF file
mtf_file = open(filename, "r")
# Reading file
for line in mtf_file:
# Processing line
line_list = line.strip().split("\t")
tf_id = line_list[0].strip()
name = line_list[1].strip()
database = line_list[2].strip()
version = int(line_list[3].strip())
gene_names = line_list[4].strip().split("+")
tf_class = line_list[5].strip()
uniprot_ids = line_list[6].strip().split(";")
data_source = line_list[7].strip() if len(line_list) > 7 else ""
self.add(MotifAnnotation(tf_id, name, database, version, gene_names, tf_class, uniprot_ids, data_source))
# Termination
mtf_file.close()
def write_network(self, targets, out_path, threshold=1):
"""
If enrichment information has been loaded before (via read_enrichment), this function creates
a cytoscape-compatible network into the output folder.
*Keyword arguments:*
- targets -- Gene targets.
- out_path -- Output path.
- threshold -- Threshold for motif acceptance.
"""
self.write_enrichment(
out_path + "/pvalue_table_" + str(threshold * 100) + ".txt",
threshold)
out_path = npath(out_path)
_, genes_motifs = self.get_mappings(key_type="gene_names")
net_pairs = {}
net_tfs = {}
all_pairs = set()
all_tfs = set()
all_genes = set()
if targets:
filter_targets = True
else:
filter_targets = False
# using genes to motif mapping to get network in all conditions
for net_name in self.networks:
net = self.networks[net_name]
pairs = set()
tfs = set()
net_pairs[net_name] = pairs
net_tfs[net_name] = tfs
for tf in genes_motifs:
motifs = genes_motifs[tf]
for m in motifs:
if m in net:
for target in net[m]:
if not filter_targets or (target in targets):
pairs.add((tf, target))
tfs.add(tf)
all_genes.add(tf)
all_genes.add(target)
else:
print("motif not in network: " + m + " " + str(tf) +
" ")
all_pairs = all_pairs.union(pairs)
all_tfs = all_tfs.union(tfs)
# printing out network
for net_name, pairs_aux in net_pairs.items():
f = open(out_path + "/" + net_name + "_targets.txt", "w")
for pair in all_pairs:
# check if pair is active in the network
if pair in pairs_aux:
f.write(pair[0] + "\t" + pair[1] + "\tactive\n")
else:
f.write(pair[0] + "\t" + pair[1] + "\tinactive\n")
f.close()
f = open(out_path + "/" + net_name + "_genes.txt", "w")
for gene in all_genes:
# check if gene is tf active in network
if gene in net_tfs[net_name]:
f.write(gene + "\ttf_active\n")
elif gene in all_tfs:
f.write(gene + "\ttf_inactive\n")
else:
f.write(gene + "\ttarget\n")
f.close()
def handle_input():
parser = HelpfulOptionParser(usage=__doc__)
parser.add_option("-n", "--name", default=None, dest="name", type="string",
help="Experiment's name and prefix for all files that are created.")
parser.add_option("-m", "--merge", default=False, dest="merge", action="store_true",
help="Merge peaks which have a distance less than the estimated mean fragment size "
"(recommended for histone data). [default: do not merge]")
parser.add_option("--housekeeping-genes", default=None, dest="housekeeping_genes", type="str",
help="Define housekeeping genes (BED format) used for normalizing. [default: %default]")
parser.add_option("--output-dir", dest="outputdir", default=None, type="string",
help="Store files in output directory. [default: %default]")
parser.add_option("--report", dest="report", default=False, action="store_true",
help="Generate HTML report about experiment. [default: %default]")
parser.add_option("--deadzones", dest="deadzones", default=None,
help="Define blacklisted genomic regions avoided for analysis (BED format). [default: %default]")
parser.add_option("--no-correction", default=False, dest="no_correction", action="store_true",
help="Do not use multipe test correction for p-values (Benjamini/Hochberg). [default: %default]")
parser.add_option("-p", "--pvalue", dest="pcutoff", default=0.1, type="float",
help="P-value cutoff for peak detection. Call only peaks with p-value lower than cutoff. "
"[default: %default]")
parser.add_option("--exts", default=None, dest="exts", type="str", action='callback', callback=_callback_list,
help="Read's extension size for BAM files (comma separated list for each BAM file in config "
"file). If option is not chosen, estimate extension sizes. [default: %default]")
parser.add_option("--factors-inputs", default=None, dest="factors_inputs", type="str", action="callback",
callback=_callback_list_float,
help="Normalization factors for input-DNA (comma separated list for each BAM file in config "
"file). If option is not chosen, estimate factors. [default: %default]")
parser.add_option("--scaling-factors", default=None, dest="scaling_factors_ip", type="str", action='callback',
callback=_callback_list_float,
help="Scaling factor for each BAM file (not control input-DNA) as comma separated list for "
"each BAM file in config file. If option is not chosen, follow normalization strategy "
"(TMM or HK approach) [default: %default]")
parser.add_option("--save-input", dest="save_input", default=False, action="store_true",
help="Save input-DNA file if available. [default: %default]")
parser.add_option("--version", dest="version", default=False, action="store_true",
help="Show script's version.")
group = OptionGroup(parser, "Advanced options")
group.add_option("--regions", dest="regions", default=None, type="string",
help="Define regions (BED format) to restrict the analysis, that is, where to train the HMM and "
"search for DPs. It is faster, but less precise.")
group.add_option("-b", "--binsize", dest="binsize", default=100, type="int",
help="Size of underlying bins for creating the signal. [default: %default]")
group.add_option("-s", "--step", dest="stepsize", default=50, type="int",
help="Stepsize with which the window consecutively slides across the genome to create the "
"signal. [default: %default]")
group.add_option("--debug", default=False, dest="debug", action="store_true",
help="Output debug information. Warning: space consuming! [default: %default]")
group.add_option("--no-gc-content", dest="no_gc_content", default=False, action="store_true",
help="Do not normalize towards GC content. [default: %default]")
group.add_option("--norm-regions", default=None, dest="norm_regions", type="str",
help="Restrict normalization to particular regions (BED format). [default: %default]")
group.add_option("-f", "--foldchange", dest="foldchange", default=1.6, type="float",
help="Fold change parameter to define training set (t_1, see paper). [default: %default]")
group.add_option("-t", "--threshold", dest="threshold", default=95, type="float",
help="Minimum signal support for differential peaks to define training set as percentage "
"(t_2, see paper). [default: %default]")
group.add_option("--size", dest="size_ts", default=10000, type="int",
help="Number of bins the HMM's training set constists of. [default: %default]")
group.add_option("--par", dest="par", default=1, type="int",
help="Percentile for p-value postprocessing filter. [default: %default]")
group.add_option("--poisson", default=False, dest="poisson", action="store_true",
help="Use binomial distribution as emmission. [default: %default]")
group.add_option("--single-strand", default=False, dest="singlestrand", action="store_true",
help="Allow single strand BAM file as input. [default: %default]")
group.add_option("--m_threshold", default=80, dest="m_threshold", type="int",
help="Define the M threshold of percentile for training TMM. [default: %default]")
group.add_option("--a_threshold", default=95, dest="a_threshold", type="int",
help="Define the A threshold of percentile for training TMM. [default: %default]")
group.add_option("--rmdup", default=False, dest="rmdup", action="store_true",
help="Remove the duplicate reads [default: %default]")
parser.add_option_group(group)
(options, args) = parser.parse_args()
options.save_wig = False
options.exts_inputs = None
options.verbose = False
options.hmm_free_para = False
if options.version:
print("")
print(__version__)
sys.exit()
if len(args) != 1:
parser.error("Please give config file")
config_path = npath(args[0])
if not isfile(config_path):
parser.error("Config file %s does not exist!" % config_path)
bamfiles, genome, chrom_sizes, inputs, dims = input_parser(config_path)
if not genome:
options.no_gc_content = True
if options.exts and len(options.exts) != len(bamfiles):
parser.error("Number of Extension Sizes must equal number of bamfiles")
if options.exts_inputs and len(options.exts_inputs) != len(inputs):
parser.error("Number of Input Extension Sizes must equal number of input bamfiles")
if options.scaling_factors_ip and len(options.scaling_factors_ip) != len(bamfiles):
parser.error("Number of scaling factors for IP must equal number of bamfiles")
for bamfile in bamfiles:
if not isfile(bamfile):
parser.error("BAM file %s does not exist!" % bamfile)
if not inputs and options.factors_inputs:
print("As no input-DNA, do not use input-DNA factors", file=sys.stderr)
options.factors_inputs = None
if options.factors_inputs and len(options.factors_inputs) != len(bamfiles):
parser.error("factors for input-DNA must equal number of BAM files!")
if inputs:
for bamfile in inputs:
if not isfile(bamfile):
parser.error("BAM file %s does not exist!" % bamfile)
if options.regions:
if not isfile(options.regions):
parser.error("Region file %s does not exist!" % options.regions)
if genome and not isfile(genome):
parser.error("Genome file %s does not exist!" % genome)
if options.name is None:
d = str(datetime.now()).replace("-", "_").replace(":", "_").replace(" ", "_").replace(".", "_").split("_")
options.name = "THOR-exp" + "-" + "_".join(d[:len(d) - 1])
if not which("wigToBigWig") or not which("bedGraphToBigWig") or not which("bigWigMerge"):
print("Warning: wigToBigWig, bigWigMerge or bedGraphToBigWig not found! Signal will not be stored!",
file=sys.stderr)
if options.outputdir:
options.outputdir = npath(options.outputdir)
if isdir(options.outputdir) and sum(
map(lambda x: x.startswith(options.name), os.listdir(options.outputdir))) > 0:
parser.error("Output directory exists and contains files with names starting with your chosen experiment "
"name! Do nothing to prevent file overwriting!")
if not exists(options.outputdir):
os.mkdir(options.outputdir)
else:
options.outputdir = os.getcwd()
options.name = join(options.outputdir, options.name)
if isdir(join(options.outputdir, 'report_'+basename(options.name))):
parser.error("Folder 'report_"+basename(options.name)+"' already exits in output directory!"
"Do nothing to prevent file overwriting! "
"Please rename report folder or change working directory of THOR with the option --output-dir")
if options.report:
os.mkdir(join(options.outputdir, 'report_'+basename(options.name)+"/"))
os.mkdir(join(options.outputdir, 'report_'+basename(options.name), 'pics/'))
os.mkdir(join(options.outputdir, 'report_'+basename(options.name), 'pics/data/'))
global FOLDER_REPORT
global FOLDER_REPORT_PICS
global FOLDER_REPORT_DATA
global OUTPUTDIR
global NAME
FOLDER_REPORT = join(options.outputdir, 'report_'+basename(options.name)+"/")
FOLDER_REPORT_PICS = join(options.outputdir, 'report_'+basename(options.name), 'pics/')
FOLDER_REPORT_DATA = join(options.outputdir, 'report_'+basename(options.name), 'pics/data/')
OUTPUTDIR = options.outputdir
NAME = options.name
if not inputs:
print("Warning: Do not compute GC-content, as there is no input file", file=sys.stderr)
if not genome:
print("Warning: Do not compute GC-content, as there is no genome file", file=sys.stderr)
if options.exts is None:
options.exts = []
if options.exts_inputs is None:
options.exts_inputs = []
return options, bamfiles, genome, chrom_sizes, dims, inputs
def write_network(self, targets, out_path, threshold=1):
"""
If enrichment information has been loaded before (via read_enrichment), this function creates
a cytoscape-compatible network into the output folder.
*Keyword arguments:*
- targets -- Gene targets.
- out_path -- Output path.
- threshold -- Threshold for motif acceptance.
"""
self.write_enrichment(out_path + "/pvalue_table_" + str(threshold * 100) + ".txt", threshold)
out_path = npath(out_path)
_, genes_motifs = self.get_mappings(key_type="gene_names")
net_pairs = {}
net_tfs = {}
all_pairs = set()
all_tfs = set()
all_genes = set()
if targets:
filter_targets = True
else:
filter_targets = False
# using genes to motif mapping to get network in all conditions
for net_name in self.networks:
net = self.networks[net_name]
pairs = set()
tfs = set()
net_pairs[net_name] = pairs
net_tfs[net_name] = tfs
for tf in genes_motifs:
motifs = genes_motifs[tf]
for m in motifs:
if m in net:
for target in net[m]:
if not filter_targets or (target in targets):
pairs.add((tf, target))
tfs.add(tf)
all_genes.add(tf)
all_genes.add(target)
else:
print("motif not in network: " + m + " " + str(tf) + " ")
all_pairs = all_pairs.union(pairs)
all_tfs = all_tfs.union(tfs)
# printing out network
for net_name, pairs_aux in net_pairs.items():
f = open(out_path + "/" + net_name + "_targets.txt", "w")
for pair in all_pairs:
# check if pair is active in the network
if pair in pairs_aux:
f.write(pair[0] + "\t" + pair[1] + "\tactive\n")
else:
f.write(pair[0] + "\t" + pair[1] + "\tinactive\n")
f.close()
f = open(out_path + "/" + net_name + "_genes.txt", "w")
for gene in all_genes:
# check if gene is tf active in network
if gene in net_tfs[net_name]:
f.write(gene + "\ttf_active\n")
elif gene in all_tfs:
f.write(gene + "\ttf_inactive\n")
else:
f.write(gene + "\ttarget\n")
f.close()
parser.add_argument('-f',
'--input-format',
choices=['jaspar-2014', 'jaspar-2016', 'hocomoco-pcm'],
type=str,
required=True,
help='format of the input file')
parser.add_argument('-o',
'--output-folder',
type=str,
required=True,
help='name of output Folder')
args = parser.parse_args()
# read the input file
with open(npath(args.input_file), "r") as f:
content = f.readlines()
n_lines = len(content)
output_folder = npath(args.output_folder)
# make output directory path, if it doesn't exist
os.makedirs(output_folder)
###################################################################################################
# JASPAR 2014
###################################################################################################
if args.input_format == "jaspar-2014":
for i in range(n_lines / 5):
def handle_input():
parser = HelpfulOptionParser(usage=__doc__)
parser.add_option(
"-n",
"--name",
default=None,
dest="name",
type="string",
help="Experiment's name and prefix for all files that are created.")
parser.add_option(
"-m",
"--merge",
default=False,
dest="merge",
action="store_true",
help=
"Merge peaks which have a distance less than the estimated mean fragment size "
"(recommended for histone data). [default: do not merge]")
parser.add_option(
"--housekeeping-genes",
default=None,
dest="housekeeping_genes",
type="str",
help=
"Define housekeeping genes (BED format) used for normalizing. [default: %default]"
)
parser.add_option(
"--output-dir",
dest="outputdir",
default=None,
type="string",
help="Store files in output directory. [default: %default]")
parser.add_option(
"--report",
dest="report",
default=False,
action="store_true",
help="Generate HTML report about experiment. [default: %default]")
parser.add_option(
"--deadzones",
dest="deadzones",
default=None,
help=
"Define blacklisted genomic regions avoided for analysis (BED format). [default: %default]"
)
parser.add_option(
"--no-correction",
default=False,
dest="no_correction",
action="store_true",
help=
"Do not use multipe test correction for p-values (Benjamini/Hochberg). [default: %default]"
)
parser.add_option(
"-p",
"--pvalue",
dest="pcutoff",
default=0.1,
type="float",
help=
"P-value cutoff for peak detection. Call only peaks with p-value lower than cutoff. "
"[default: %default]")
parser.add_option(
"--exts",
default=None,
dest="exts",
type="str",
action='callback',
callback=_callback_list,
help=
"Read's extension size for BAM files (comma separated list for each BAM file in config "
"file). If option is not chosen, estimate extension sizes. [default: %default]"
)
parser.add_option(
"--factors-inputs",
default=None,
dest="factors_inputs",
type="str",
action="callback",
callback=_callback_list_float,
help=
"Normalization factors for input-DNA (comma separated list for each BAM file in config "
"file). If option is not chosen, estimate factors. [default: %default]"
)
parser.add_option(
"--scaling-factors",
default=None,
dest="scaling_factors_ip",
type="str",
action='callback',
callback=_callback_list_float,
help=
"Scaling factor for each BAM file (not control input-DNA) as comma separated list for "
"each BAM file in config file. If option is not chosen, follow normalization strategy "
"(TMM or HK approach) [default: %default]")
parser.add_option(
"--save-input",
dest="save_input",
default=False,
action="store_true",
help="Save input-DNA file if available. [default: %default]")
parser.add_option("--version",
dest="version",
default=False,
action="store_true",
help="Show script's version.")
group = OptionGroup(parser, "Advanced options")
group.add_option(
"--regions",
dest="regions",
default=None,
type="string",
help=
"Define regions (BED format) to restrict the analysis, that is, where to train the HMM and "
"search for DPs. It is faster, but less precise.")
group.add_option(
"-b",
"--binsize",
dest="binsize",
default=100,
type="int",
help=
"Size of underlying bins for creating the signal. [default: %default]")
group.add_option(
"-s",
"--step",
dest="stepsize",
default=50,
type="int",
help=
"Stepsize with which the window consecutively slides across the genome to create the "
"signal. [default: %default]")
group.add_option(
"--debug",
default=False,
dest="debug",
action="store_true",
help=
"Output debug information. Warning: space consuming! [default: %default]"
)
group.add_option(
"--no-gc-content",
dest="no_gc_content",
default=False,
action="store_true",
help="Do not normalize towards GC content. [default: %default]")
group.add_option(
"--norm-regions",
default=None,
dest="norm_regions",
type="str",
help=
"Restrict normalization to particular regions (BED format). [default: %default]"
)
group.add_option(
"-f",
"--foldchange",
dest="foldchange",
default=1.6,
type="float",
help=
"Fold change parameter to define training set (t_1, see paper). [default: %default]"
)
group.add_option(
"-t",
"--threshold",
dest="threshold",
default=95,
type="float",
help=
"Minimum signal support for differential peaks to define training set as percentage "
"(t_2, see paper). [default: %default]")
group.add_option(
"--size",
dest="size_ts",
default=10000,
type="int",
help=
"Number of bins the HMM's training set constists of. [default: %default]"
)
group.add_option(
"--par",
dest="par",
default=1,
type="int",
help="Percentile for p-value postprocessing filter. [default: %default]"
)
group.add_option(
"--poisson",
default=False,
dest="poisson",
action="store_true",
help="Use binomial distribution as emmission. [default: %default]")
group.add_option(
"--single-strand",
default=False,
dest="singlestrand",
action="store_true",
help="Allow single strand BAM file as input. [default: %default]")
group.add_option(
"--m_threshold",
default=80,
dest="m_threshold",
type="int",
help=
"Define the M threshold of percentile for training TMM. [default: %default]"
)
group.add_option(
"--a_threshold",
default=95,
dest="a_threshold",
type="int",
help=
"Define the A threshold of percentile for training TMM. [default: %default]"
)
group.add_option("--rmdup",
default=False,
dest="rmdup",
action="store_true",
help="Remove the duplicate reads [default: %default]")
parser.add_option_group(group)
(options, args) = parser.parse_args()
options.save_wig = False
options.exts_inputs = None
options.verbose = False
options.hmm_free_para = False
if options.version:
print("")
print(__version__)
sys.exit()
if len(args) != 1:
parser.error("Please give config file")
config_path = npath(args[0])
if not isfile(config_path):
parser.error("Config file %s does not exist!" % config_path)
bamfiles, genome, chrom_sizes, inputs, dims = input_parser(config_path)
if not genome:
options.no_gc_content = True
if options.exts and len(options.exts) != len(bamfiles):
parser.error("Number of Extension Sizes must equal number of bamfiles")
if options.exts_inputs and len(options.exts_inputs) != len(inputs):
parser.error(
"Number of Input Extension Sizes must equal number of input bamfiles"
)
if options.scaling_factors_ip and len(
options.scaling_factors_ip) != len(bamfiles):
parser.error(
"Number of scaling factors for IP must equal number of bamfiles")
for bamfile in bamfiles:
if not isfile(bamfile):
parser.error("BAM file %s does not exist!" % bamfile)
if not inputs and options.factors_inputs:
print("As no input-DNA, do not use input-DNA factors", file=sys.stderr)
options.factors_inputs = None
if options.factors_inputs and len(options.factors_inputs) != len(bamfiles):
parser.error("factors for input-DNA must equal number of BAM files!")
if inputs:
for bamfile in inputs:
if not isfile(bamfile):
parser.error("BAM file %s does not exist!" % bamfile)
if options.regions:
if not isfile(options.regions):
parser.error("Region file %s does not exist!" % options.regions)
if genome and not isfile(genome):
parser.error("Genome file %s does not exist!" % genome)
if options.name is None:
d = str(datetime.now()).replace("-", "_").replace(":", "_").replace(
" ", "_").replace(".", "_").split("_")
options.name = "THOR-exp" + "-" + "_".join(d[:len(d) - 1])
if not which("wigToBigWig") or not which("bedGraphToBigWig") or not which(
"bigWigMerge"):
print(
"Warning: wigToBigWig, bigWigMerge or bedGraphToBigWig not found! Signal will not be stored!",
file=sys.stderr)
if options.outputdir:
options.outputdir = npath(options.outputdir)
if isdir(options.outputdir) and sum(
map(lambda x: x.startswith(options.name),
os.listdir(options.outputdir))) > 0:
parser.error(
"Output directory exists and contains files with names starting with your chosen experiment "
"name! Do nothing to prevent file overwriting!")
if not exists(options.outputdir):
os.mkdir(options.outputdir)
else:
options.outputdir = os.getcwd()
options.name = join(options.outputdir, options.name)
if isdir(join(options.outputdir, 'report_' + basename(options.name))):
parser.error(
"Folder 'report_" + basename(options.name) +
"' already exits in output directory!"
"Do nothing to prevent file overwriting! "
"Please rename report folder or change working directory of THOR with the option --output-dir"
)
if options.report:
os.mkdir(
join(options.outputdir, 'report_' + basename(options.name) + "/"))
os.mkdir(
join(options.outputdir, 'report_' + basename(options.name),
'pics/'))
os.mkdir(
join(options.outputdir, 'report_' + basename(options.name),
'pics/data/'))
global FOLDER_REPORT
global FOLDER_REPORT_PICS
global FOLDER_REPORT_DATA
global OUTPUTDIR
global NAME
FOLDER_REPORT = join(options.outputdir,
'report_' + basename(options.name) + "/")
FOLDER_REPORT_PICS = join(options.outputdir,
'report_' + basename(options.name), 'pics/')
FOLDER_REPORT_DATA = join(options.outputdir,
'report_' + basename(options.name), 'pics/data/')
OUTPUTDIR = options.outputdir
NAME = options.name
if not inputs:
print("Warning: Do not compute GC-content, as there is no input file",
file=sys.stderr)
if not genome:
print("Warning: Do not compute GC-content, as there is no genome file",
file=sys.stderr)
if options.exts is None:
options.exts = []
if options.exts_inputs is None:
options.exts_inputs = []
return options, bamfiles, genome, chrom_sizes, dims, inputs
