def init_gerp_dic(motif_dic):
'''
Read pickle containing motif dic, then
extract all the chromosomes and genomic coords
that will need to be read in RS score.
motif dic is expected to be of form:
{miso_event: {region: {subkey1: subval1 ...}}}
Returns:
gerp_dic of form:
{chr1: {coord1: None, coord1: None...}
chrY: {...}}
'''
# Define subkey constants. Should match get_dic_subkeys()
genomic_coord_str = 'genomic_coordinate'
# Init chromosome list
chr_list = get_chr_list()
# Init gerp_dic
gerp_dic = {}
# Create empty subdics, by chromosmoe
for chromosome in chr_list:
gerp_dic[chromosome] = {}
# Iterate through the different subdics, retrieving genomic coords
# store genomic coords to gerp_dic
coord_count = 0
for miso_event in motif_dic:
for region in motif_dic[miso_event]:
# get colon separated chr_start_end value.
chr_start_end_list = \
motif_dic[miso_event][region][genomic_coord_str]
# loop through list of coords (mostly length 1)
for chr_start_end in chr_start_end_list:
if chr_start_end is not None:
# Split by colon
chr_start_end_split = chr_start_end.split(':')
# Get chromo, start, ends separately.
# Start is inclusive in RS score file.
# End is not inclusive (so we subtract by 1).
chromo = chr_start_end_split[0]
start = int(chr_start_end_split[1])
end = int(chr_start_end_split[2])
# Update gerp_dic containing empty dics
# with coordinates as keys, from start to end
# Note: end is not inclusive.
for coordinate in range(start, end):
gerp_dic[chromo][coordinate] = None
coord_count += 1
print '%s coordinates stored into gerp_dic.' %coord_count
return gerp_dic
def init_gerp_dic(motif_dic):
'''
Read pickle containing motif dic, then
extract all the chromosomes and genomic coords
that will need to be read in RS score.
motif dic is expected to be of form:
{miso_event: {region: {subkey1: subval1 ...}}}
Returns:
gerp_dic of form:
{chr1: {coord1: None, coord1: None...}
chrY: {...}}
'''
# Define subkey constants. Should match get_dic_subkeys()
genomic_coord_str = 'genomic_coordinate'
# Init chromosome list
chr_list = get_chr_list()
# Init gerp_dic
gerp_dic = {}
# Create empty subdics, by chromosmoe
for chromosome in chr_list:
gerp_dic[chromosome] = {}
# Iterate through the different subdics, retrieving genomic coords
# store genomic coords to gerp_dic
coord_count = 0
for miso_event in motif_dic:
for region in motif_dic[miso_event]:
# get colon separated chr_start_end value.
chr_start_end_list = \
motif_dic[miso_event][region][genomic_coord_str]
# loop through list of coords (mostly length 1)
for chr_start_end in chr_start_end_list:
if chr_start_end is not None:
# Split by colon
chr_start_end_split = chr_start_end.split(':')
# Get chromo, start, ends separately.
# Start is inclusive in RS score file.
# End is not inclusive (so we subtract by 1).
chromo = chr_start_end_split[0]
start = int(chr_start_end_split[1])
end = int(chr_start_end_split[2])
# Update gerp_dic containing empty dics
# with coordinates as keys, from start to end
# Note: end is not inclusive.
for coordinate in range(start, end):
gerp_dic[chromo][coordinate] = None
coord_count += 1
print '%s coordinates stored into gerp_dic.' % coord_count
return gerp_dic
def main():
usage = 'usage: %prog pickle_filepath gerp_directory output_file\n'\
'Two args must be specified in commandline: \n'\
'1) Path to pickle from summarize_meme_results.py\n'\
'2) Directory containing GERP RS score text files by chromosome\n'\
'3) Output file to which results will be written.\n'
parser = OptionParser(usage=usage)
parser.add_option('-p', '--gerp_pickle_fname', dest='gerp_pickle_fname',
default='gerp_pickle.pkl',
help='gerp scores pickle filename.\n'\
'Default gerp_pickle.pkl')
parser.add_option('-l', '--presaved_gerp_dic_path', dest='gerp_presaved_path',
default=None,
help='If a gerp dic has been presaved, use this flag to'\
'indicate the file path to directly open the gerp file. '\
'Reduces need for multiprocessing.')
parser.add_option('-o', '--updated_dic_fname', dest='updated_dic_fname',
default='meme_summary.gerp_updated.pkl',
help='Filename to updated meme_summary.pkl'
' with gerp scores,\n'\
'Default "meme_summary.gerp_updated.pkl"')
parser.add_option('-t', '--conservation_type', dest='cons_type',
default='rs',
help='Either "rs" or "phastcons"')
parser.add_option('-m', '--multiprocessing', dest='multiprocessing',
default=True,
help='True or False. Sets multiprocessing or not.')
(options, args) = parser.parse_args()
if len(args) < 2:
print 'Incorrect number of parameters specified.'
print usage
sys.exit()
motif_pickle_path = args[0]
gerp_dir = args[1]
output_path = args[2]
# parse options
gerp_pickle_fname = options.gerp_pickle_fname
gerp_presaved_pkl_path = options.gerp_presaved_path
updated_dic_fname = options.updated_dic_fname
cons_type = options.cons_type
if cons_type not in ['rs', 'phastcons']:
print 'Conservation type must be "rs" or "phastcons". %s found.' \
%cons_type
sys.exit()
# Load motif dic, obtained from summarize_meme_results
pickle_file = open(motif_pickle_path, 'rb')
motif_dic = pickle.load(pickle_file)
pickle_file.close()
# Get chromosome list
chr_list = get_chr_list()
# Create dic containing chromosomes and genomic coordinates
# relevant to our motifs by reading pickled dictionary.
gerp_dic = init_gerp_dic(motif_dic)
# Multithread only if presaved_gerp_dic_path flag is None.
if gerp_presaved_pkl_path == None:
# BEGIN: MULTITHREADING
print 'Beginning multiprocessing.'
q = Queue()
process_list = []
# For each chromosome, open relevant gerp file and retrieve
# RS scores associated with the coordinates.
for chromosome in chr_list:
p = Process(target=add_rs_scores_to_gerp_dic,
args=(gerp_dic, chromosome, gerp_dir, q, cons_type))
process_list.append(p)
p.start()
print 'Calculating RS scores for %s' %chromosome
print 'Finsihed calculating RS scores.'
for _ in chr_list:
# Update dic for every process that started.
# the actual chromosoem doesn't matter. It's the
# number of iterations that matter.
(gerp_dic_chromo, chromo) = q.get()
print 'Updating dictionary for chromo: %s' %chromo
# Find which chromosome this came from by looking at key
gerp_dic[chromo].update(gerp_dic_chromo)
# Wait for all threads to be done before continuing.
for p in process_list:
p.join()
# END: MULTITHREADING
print 'Done multiprocessing.'
# save gerp dic as pickle
gerp_pickle_dir = os.path.dirname(motif_pickle_path)
gerp_pickle_fpath = os.path.join(gerp_pickle_dir, gerp_pickle_fname)
with open(gerp_pickle_fpath, 'wb') as gerp_pickle_file:
pickle.dump(gerp_dic, gerp_pickle_file, -1)
print 'Saved pickle object to: %s' %gerp_pickle_fpath
else:
# Try to open pickle path from options.gerp_path
with open(gerp_presaved_pkl_path, 'rb') as presaved_pkl:
pickle.load(presaved_pkl)
print 'Loaded presaved pickle from %s' %gerp_presaved_pkl_path
# Update motif_dic with gerp_dic
motif_dic = update_motif_dic_with_gerp_dic(motif_dic, gerp_dic)
# Save updated motif dic to pickle
updated_dic_dir = os.path.dirname(motif_pickle_path)
updated_dic_path = os.path.join(updated_dic_dir, updated_dic_fname)
with open(updated_dic_path, 'wb') as updated_dic_file:
pickle.dump(motif_dic, updated_dic_file, -1)
print 'Updated dic saved to: %s' %updated_dic_path
# Write updated gerp_dic to file
# add GERP scores as a subkey in subkey_list
subkeys_list = get_dic_subkeys()
rs_score_subkey = get_rs_score_subkey()
subkeys_list.append(rs_score_subkey)
# Write updated motif dic to file
write_outdic_to_file(motif_dic, output_path, subkeys_list)
def main():
usage = 'usage: %prog pickle_filepath gerp_directory output_file\n'\
'Two args must be specified in commandline: \n'\
'1) Path to pickle from summarize_meme_results.py\n'\
'2) Directory containing GERP RS score text files by chromosome\n'\
'3) Output file to which results will be written.\n'
parser = OptionParser(usage=usage)
parser.add_option('-p', '--gerp_pickle_fname', dest='gerp_pickle_fname',
default='gerp_pickle.pkl',
help='gerp scores pickle filename.\n'\
'Default gerp_pickle.pkl')
parser.add_option('-l', '--presaved_gerp_dic_path', dest='gerp_presaved_path',
default=None,
help='If a gerp dic has been presaved, use this flag to'\
'indicate the file path to directly open the gerp file. '\
'Reduces need for multiprocessing.')
parser.add_option('-o', '--updated_dic_fname', dest='updated_dic_fname',
default='meme_summary.gerp_updated.pkl',
help='Filename to updated meme_summary.pkl'
' with gerp scores,\n'\
'Default "meme_summary.gerp_updated.pkl"')
parser.add_option('-t',
'--conservation_type',
dest='cons_type',
default='rs',
help='Either "rs" or "phastcons"')
parser.add_option('-m',
'--multiprocessing',
dest='multiprocessing',
default=True,
help='True or False. Sets multiprocessing or not.')
(options, args) = parser.parse_args()
if len(args) < 2:
print 'Incorrect number of parameters specified.'
print usage
sys.exit()
motif_pickle_path = args[0]
gerp_dir = args[1]
output_path = args[2]
# parse options
gerp_pickle_fname = options.gerp_pickle_fname
gerp_presaved_pkl_path = options.gerp_presaved_path
updated_dic_fname = options.updated_dic_fname
cons_type = options.cons_type
if cons_type not in ['rs', 'phastcons']:
print 'Conservation type must be "rs" or "phastcons". %s found.' \
%cons_type
sys.exit()
# Load motif dic, obtained from summarize_meme_results
pickle_file = open(motif_pickle_path, 'rb')
motif_dic = pickle.load(pickle_file)
pickle_file.close()
# Get chromosome list
chr_list = get_chr_list()
# Create dic containing chromosomes and genomic coordinates
# relevant to our motifs by reading pickled dictionary.
gerp_dic = init_gerp_dic(motif_dic)
# Multithread only if presaved_gerp_dic_path flag is None.
if gerp_presaved_pkl_path == None:
# BEGIN: MULTITHREADING
print 'Beginning multiprocessing.'
q = Queue()
process_list = []
# For each chromosome, open relevant gerp file and retrieve
# RS scores associated with the coordinates.
for chromosome in chr_list:
p = Process(target=add_rs_scores_to_gerp_dic,
args=(gerp_dic, chromosome, gerp_dir, q, cons_type))
process_list.append(p)
p.start()
print 'Calculating RS scores for %s' % chromosome
print 'Finsihed calculating RS scores.'
for _ in chr_list:
# Update dic for every process that started.
# the actual chromosoem doesn't matter. It's the
# number of iterations that matter.
(gerp_dic_chromo, chromo) = q.get()
print 'Updating dictionary for chromo: %s' % chromo
# Find which chromosome this came from by looking at key
gerp_dic[chromo].update(gerp_dic_chromo)
# Wait for all threads to be done before continuing.
for p in process_list:
p.join()
# END: MULTITHREADING
print 'Done multiprocessing.'
# save gerp dic as pickle
gerp_pickle_dir = os.path.dirname(motif_pickle_path)
gerp_pickle_fpath = os.path.join(gerp_pickle_dir, gerp_pickle_fname)
with open(gerp_pickle_fpath, 'wb') as gerp_pickle_file:
pickle.dump(gerp_dic, gerp_pickle_file, -1)
print 'Saved pickle object to: %s' % gerp_pickle_fpath
else:
# Try to open pickle path from options.gerp_path
with open(gerp_presaved_pkl_path, 'rb') as presaved_pkl:
pickle.load(presaved_pkl)
print 'Loaded presaved pickle from %s' % gerp_presaved_pkl_path
# Update motif_dic with gerp_dic
motif_dic = update_motif_dic_with_gerp_dic(motif_dic, gerp_dic)
# Save updated motif dic to pickle
updated_dic_dir = os.path.dirname(motif_pickle_path)
updated_dic_path = os.path.join(updated_dic_dir, updated_dic_fname)
with open(updated_dic_path, 'wb') as updated_dic_file:
pickle.dump(motif_dic, updated_dic_file, -1)
print 'Updated dic saved to: %s' % updated_dic_path
# Write updated gerp_dic to file
# add GERP scores as a subkey in subkey_list
subkeys_list = get_dic_subkeys()
rs_score_subkey = get_rs_score_subkey()
subkeys_list.append(rs_score_subkey)
# Write updated motif dic to file
write_outdic_to_file(motif_dic, output_path, subkeys_list)
