def main():
main_dir = sys.argv[1]
sample_dir_fullpath = sys.argv[2]
output_path = sys.argv[3]
# Define constants
chr_str = 'chr'
# Create list of chromosome names corresponding to folders within sample dir
chr_list = [''.join([chr_str, str(c)]) for c in range(1, 23) + ['X', 'Y']]
# Create list of sample directory names.
sample_dir_names_list = get_sample_names_from_file(sample_dir_fullpath)
# Subset list for only those that contain miso outputs.
sample_dir_names_list = check_if_empty_dir(main_dir, sample_dir_names_list, chr_list)
# sample_dir_names_list = sample_dir_names_csv.split(',')
# Run on multiple threads.
for chromo in chr_list:
print('Sending %s job to core...' %chromo)
Process(target=get_fnames_consolidate_miso,
args=(main_dir, sample_dir_names_list, chromo,
output_path)).start()
def main():
# Get parse options.
usage = 'usage: %prog [options] t_test_file samp1_file samp2_file write_output_file'
parser = OptionParser(usage=usage)
(_, args) = parser.parse_args()
if len(args) < 4:
print('Following must be specified in command line:\n'\
't_test_file (filtered, preferably)\n'\
'samp1_file (group1 samps: should match sample names in t_test_file\n'\
'samp2_file (group2 samps: should match samp names in t_test_file\n'\
'output_file')
sys.exit()
# Define args to variables
t_test_file = args[0]
samp1_file = args[1]
samp2_file = args[2]
output_file = args[3]
# Define colnames in ttest output
gsymbol_str = 'gsymbol'
sample_name_str = 'sample_name'
psi_median_str = 'psi_median'
# Get sample names from textfile.
group_1_samples = \
group_miso_utils.get_sample_names_from_file(samp1_file)
group_2_samples = \
group_miso_utils.get_sample_names_from_file(samp2_file)
all_samples = group_1_samples + group_2_samples
# Init my write file
writefile = open(output_file, 'wb')
mywriter = csv.writer(writefile, delimiter='\t')
# Write row of columns
# first column reserved for genenames therfore we use ['gsymbol']
write_header = [gsymbol_str] + all_samples
mywriter.writerow(write_header)
# Read t_test_file
with open(t_test_file, 'rb') as readfile:
myreader = csv.reader(readfile, delimiter='\t')
header = myreader.next()
writecount = 0
gsymbol_list = [] # keep track of genes, avoid duplicates.
for row in myreader:
'''
For each row, grab sample names and their PSI median values.
Many rows will not have PSI medians due to low read count.
Replace missing PSI medians with NAs.
For each iteration, initialize a dictionary of form:
{samp: 'NA'}
fill in psi meds into a dictionary of form:
{samp: psi_med}
'''
# Initialize dictionary
row_dic = init_samp_dic(all_samples + [gsymbol_str])
# Get gsymbol, row samps (as list), psi meds (as list)
gsymbol = get_row_info(row, header, gsymbol_str, csv=False)
if gsymbol == 'NA':
gsymbol = 'no_name'
# Check if gsymbol is already in gsymbol list, modify genename if so.
gsymbol_checked = check_not_duplicate(gsymbol, gsymbol_list)
gsymbol_list.append(gsymbol)
row_samps = get_row_info(row, header, sample_name_str, csv=True)
row_psi_meds = get_row_info(row, header, psi_median_str, csv=True)
# Fill in rowsamps, psimeds, gsymbol into our initialized row_dic
row_dic = fill_dic(row_dic,
row_samps + [gsymbol_str],
row_psi_meds + [gsymbol_checked])
# Write row_dic and gsymbol to file.
write_row_dic_to_file(row_dic,
write_header,
mywriter)
writecount += 1
print('%s rows written to: %s' %(writecount, output_file))
writefile.close()
def main():
# Get parse options.
usage = 'usage: %prog [options] t_test_file samp1_file samp2_file write_output_file'
parser = OptionParser(usage=usage)
(_, args) = parser.parse_args()
if len(args) < 4:
print('Following must be specified in command line:\n'\
't_test_file (filtered, preferably)\n'\
'samp1_file (group1 samps: should match sample names in t_test_file\n'\
'samp2_file (group2 samps: should match samp names in t_test_file\n'\
'output_file')
sys.exit()
# Define args to variables
t_test_file = args[0]
samp1_file = args[1]
samp2_file = args[2]
output_file = args[3]
# Define colnames in ttest output
gsymbol_str = 'gsymbol'
sample_name_str = 'sample_name'
psi_median_str = 'psi_median'
# Get sample names from textfile.
group_1_samples = \
group_miso_utils.get_sample_names_from_file(samp1_file)
group_2_samples = \
group_miso_utils.get_sample_names_from_file(samp2_file)
all_samples = group_1_samples + group_2_samples
# Init my write file
writefile = open(output_file, 'wb')
mywriter = csv.writer(writefile, delimiter='\t')
# Write row of columns
# first column reserved for genenames therfore we use ['gsymbol']
write_header = [gsymbol_str] + all_samples
mywriter.writerow(write_header)
# Read t_test_file
with open(t_test_file, 'rb') as readfile:
myreader = csv.reader(readfile, delimiter='\t')
header = myreader.next()
writecount = 0
gsymbol_list = [] # keep track of genes, avoid duplicates.
for row in myreader:
'''
For each row, grab sample names and their PSI median values.
Many rows will not have PSI medians due to low read count.
Replace missing PSI medians with NAs.
For each iteration, initialize a dictionary of form:
{samp: 'NA'}
fill in psi meds into a dictionary of form:
{samp: psi_med}
'''
# Initialize dictionary
row_dic = init_samp_dic(all_samples + [gsymbol_str])
# Get gsymbol, row samps (as list), psi meds (as list)
gsymbol = get_row_info(row, header, gsymbol_str, csv=False)
if gsymbol == 'NA':
gsymbol = 'no_name'
# Check if gsymbol is already in gsymbol list, modify genename if so.
gsymbol_checked = check_not_duplicate(gsymbol, gsymbol_list)
gsymbol_list.append(gsymbol)
row_samps = get_row_info(row, header, sample_name_str, csv=True)
row_psi_meds = get_row_info(row, header, psi_median_str, csv=True)
# Fill in rowsamps, psimeds, gsymbol into our initialized row_dic
row_dic = fill_dic(row_dic, row_samps + [gsymbol_str],
row_psi_meds + [gsymbol_checked])
# Write row_dic and gsymbol to file.
write_row_dic_to_file(row_dic, write_header, mywriter)
writecount += 1
print('%s rows written to: %s' % (writecount, output_file))
writefile.close()
def main():
parser = OptionParser()
parser.add_option('-1', '--group1_file', dest='group_1_samplenames_file',
help='Filename containing group 1 sample names (PCa)')
parser.add_option('-2', '--group2_file', dest='group_2_samplenames_file',
help='Filename containing group 2 sample names (NEPC)')
parser.add_option('-d', '--main_directory', dest='main_dir',
help='Main directory containing miso output results.')
parser.add_option('-o', '--output_directory', dest='output_dir',
help='Output directory of t-test results.')
parser.add_option('-O', '--output_filename', dest='output_fname',
help='Output filename of the t-test results.')
parser.add_option('-m', '--min_counts', type='int', dest='min_counts',
help='Minimum junction read counts to be considered '\
'into the t-test. Best practices says 10.')
# Parse options
(options, _) = parser.parse_args()
# Define constants from options
group_1_samplenames_file = options.group_1_samplenames_file
group_2_samplenames_file = options.group_2_samplenames_file
main_dir = options.main_dir
output_dir = options.output_dir
output_fname = options.output_fname
min_counts = options.min_counts
# Define constants
summary_fullpath = os.path.join(output_dir, output_fname)
# Get sample names from textfile.
group_1_samples = get_sample_names_from_file(group_1_samplenames_file)
group_2_samples = get_sample_names_from_file(group_2_samplenames_file)
# Create list of chromosomes.
chr_list = create_chromo_list(prefix='chr')
# chr_list = ['chr11']
# Subset list for only those that contain miso outputs.
group_1_samples = check_if_empty_dir(main_dir, group_1_samples, chr_list)
group_2_samples = check_if_empty_dir(main_dir, group_2_samples, chr_list)
# Init fnames dic
fnames_dic = {}
# Run on multiple threads.
q = Queue()
process_list = []
for chromo in chr_list:
print('Sending %s job to core...' %chromo)
p = Process(target=t_test_and_pickle,
args=(fnames_dic, chromo, output_dir,
group_1_samples, group_2_samples,
main_dir, q, min_counts))
process_list.append(p)
p.start()
for chromo in chr_list:
fnames_dic.update(q.get())
# Wait for all threads to be done before continuing.
for p in process_list:
p.join()
print('Completed %s jobs.' %len(chr_list))
# Write fnames_dic as pickle file.
pickle_filename = ''.join([output_fname, '_filenames_dic.pickle'])
fnames_savepath = os.path.join(output_dir, pickle_filename)
print('Saving filenames_dic.pickle to %s' %fnames_savepath)
pickle_path = save_dic_as_pickle(fnames_dic, fnames_savepath)
# Write information from pickle to textfile.
print('Writing information from pickle to textfile.')
# Read pickle file to get fnames_dic
fnames_dic = read_pickle(pickle_path)
# Read and write to file.
read_pickle_write_to_file(summary_fullpath, chr_list, fnames_dic,
filter_events=True)
print('Summary file saved in: %s' %summary_fullpath)
def main():
parser = OptionParser()
parser.add_option('-1',
'--group1_file',
dest='group_1_samplenames_file',
help='Filename containing group 1 sample names (PCa)')
parser.add_option('-2',
'--group2_file',
dest='group_2_samplenames_file',
help='Filename containing group 2 sample names (NEPC)')
parser.add_option('-d',
'--main_directory',
dest='main_dir',
help='Main directory containing miso output results.')
parser.add_option('-o',
'--output_directory',
dest='output_dir',
help='Output directory of t-test results.')
parser.add_option('-O',
'--output_filename',
dest='output_fname',
help='Output filename of the t-test results.')
parser.add_option('-m', '--min_counts', type='int', dest='min_counts',
help='Minimum junction read counts to be considered '\
'into the t-test. Best practices says 10.')
# Parse options
(options, _) = parser.parse_args()
# Define constants from options
group_1_samplenames_file = options.group_1_samplenames_file
group_2_samplenames_file = options.group_2_samplenames_file
main_dir = options.main_dir
output_dir = options.output_dir
output_fname = options.output_fname
min_counts = options.min_counts
# Define constants
summary_fullpath = os.path.join(output_dir, output_fname)
# Get sample names from textfile.
group_1_samples = get_sample_names_from_file(group_1_samplenames_file)
group_2_samples = get_sample_names_from_file(group_2_samplenames_file)
# Create list of chromosomes.
chr_list = create_chromo_list(prefix='chr')
# chr_list = ['chr11']
# Subset list for only those that contain miso outputs.
group_1_samples = check_if_empty_dir(main_dir, group_1_samples, chr_list)
group_2_samples = check_if_empty_dir(main_dir, group_2_samples, chr_list)
# Init fnames dic
fnames_dic = {}
# Run on multiple threads.
q = Queue()
process_list = []
for chromo in chr_list:
print('Sending %s job to core...' % chromo)
p = Process(target=t_test_and_pickle,
args=(fnames_dic, chromo, output_dir, group_1_samples,
group_2_samples, main_dir, q, min_counts))
process_list.append(p)
p.start()
for chromo in chr_list:
fnames_dic.update(q.get())
# Wait for all threads to be done before continuing.
for p in process_list:
p.join()
print('Completed %s jobs.' % len(chr_list))
# Write fnames_dic as pickle file.
pickle_filename = ''.join([output_fname, '_filenames_dic.pickle'])
fnames_savepath = os.path.join(output_dir, pickle_filename)
print('Saving filenames_dic.pickle to %s' % fnames_savepath)
pickle_path = save_dic_as_pickle(fnames_dic, fnames_savepath)
# Write information from pickle to textfile.
print('Writing information from pickle to textfile.')
# Read pickle file to get fnames_dic
fnames_dic = read_pickle(pickle_path)
# Read and write to file.
read_pickle_write_to_file(summary_fullpath,
chr_list,
fnames_dic,
filter_events=True)
print('Summary file saved in: %s' % summary_fullpath)
