conttype = sys.argv[2] # contacts.gz or oe.gz
# chr_num="12,13,14"
# conttype = "contacts.gz"
logging.basicConfig(level=logging.DEBUG)
if __name__ == '__main__': # Requiered for parallelization, at least on Windows
for conttype in [conttype]:
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "input/K562/"
output_folder = "output/K562/"
cell_type = "K562"
params = Parameters()
params.window_size = 25000 # region around contact to be binned for predictors
params.mindist = 50001 # minimum distance between contacting regions
params.maxdist = 1500000 # maximum distance between contacting regions
params.sample_size = 100 # how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient = False # set True if you want use only CTCF with orient
#params.use_only_contacts_with_CTCF = "cont_with_CTCF" # "cont_with_CTCF"
params.use_only_contacts_with_CTCF = "no"
# use this option to change proportion
# of contacts with nearest ctcf sites in training datasets
write_all_chrms_in_file = True # set True if you have train with few chromosomes. Need for writing different chromosomes in the same file
fill_empty_contacts = False
conttype = sys.argv[2]
# chr_num="12"
# conttype = "contacts.gz"
if __name__ == '__main__': #Requered for parallization, at least on Windows
#,"chr10", "chr1"]:
for conttype in [conttype]:
print("hello")
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s', datefmt='%I:%M:%S', level=logging.DEBUG)
input_folder ="/mnt/scratch/ws/psbelokopytova/202001051010polina_data/3DPredictor/input/K562/"
output_folder = "/mnt/scratch/ws/psbelokopytova/202001051010polina_data/3DPredictor/out/K562/5KB/all_predictors/"
cell_type="K562"
lengths_dict = {'chr1': 1494930, 'chr3': 609806, 'chr5': 518646, 'chr7': 682860, 'chr11': 726290, 'chr13': 115324}
params = Parameters()
params.binsize = 5000 #sequence resolution of contacts data. Use for finding of normalized coefficient file
params.window_size = params.binsize #region around contact to be binned for predictors. Usually equal to binsize
params.mindist = params.binsize*2+1 #minimum distance between contacting regions
params.maxdist = 1500000
params.sample_size = 250000 #how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient=False
params.use_only_contacts_with_CTCF = "all_cont"#"cont_with_CTCF"#"#"all_cont"#"cont_with_CTCF "
write_all_chrms_in_file=False #set True if you want write training file consisting several chromosomes
fill_empty_contacts = False #set True if you want use all contacts in region, without empty contacts
logging.getLogger(__name__).debug("Using input folder "+input_folder)
import pandas as pd
import os
if __name__ == '__main__': #Requered for parallization, at least on Windows
#,"chr10", "chr1"]:
for conttype in ["contacts.gz", "oe.gz"]:
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "input/Hepat/"
#output_folder = "D:/Users/Polina/3Dpredictor/"
output_folder = "out/Hepat/validating_chrms/"
#input_folder = "input"
params = Parameters()
params.window_size = 25000 #region around contact to be binned for predictors
#params.small_window_size = 12500 #region around contact ancors to be considered as cis
params.mindist = 50001 #minimum distance between contacting regions
#params.maxdist = params.window_size #max distance between contacting regions
params.maxdist = 1500000
#params.binsize = 20000 #when binning regions with predictors, use this binsize
params.sample_size = 25000 #how many contacts write to file
#params.conttype = "oe.gz"
params.conttype = conttype
params.max_cpus = 12
logging.getLogger(__name__).debug("Using input folder " + input_folder)
#Read contacts data
params.contacts_reader = ContactsReader()
if __name__ == '__main__':
parser = createParser()
namespace = parser.parse_args(sys.argv[1:])
RNA_seq_file = namespace.RNA_seq_file
CTCF_file = namespace.CTCF_file
CTCF_orient_file = namespace.CTCF_orient_file
chr = namespace.chr
interval_start = namespace.interval_start
interval_end = namespace.interval_end
resolution = int(namespace.resolution)
model_path = namespace.model_path
out_file = namespace.out_file
params = Parameters()
params.window_size = int(
resolution) #region around contact to be binned for predictors
params.mindist = int(
resolution) * 2 + 1 #minimum distance between contacting regions
params.maxdist = 1500000 #maximum distance between contacting regions
params.max_cpus = 1
# params.keep_only_orient=False #set True if you want use only CTCF with orient
params.multiprocessing = False
# params.write_to_file = False
# Read CTCF data
# CTCF_file format: ENCODE narrow peak
# CTCF_orient_file format: chr--start--end--name--score--strand
logging.info('create CTCF_PG')
# set path to the CTCF chip-seq file:
params.ctcf_reader = ChiPSeqReader(CTCF_file, name="CTCF")
conttype = sys.argv[2]
# chr_num="12"
# conttype = "contacts.gz"
if __name__ == '__main__': #Requered for parallization, at least on Windows
#,"chr10", "chr1"]:
for conttype in [conttype]:
print("hello")
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s', datefmt='%I:%M:%S', level=logging.DEBUG)
input_folder ="/mnt/scratch/ws/psbelokopytova/201907031108polinaB/3DPredictor/input/NPC/"
#output_folder = "D:/Users/Polina/3Dpredictor/"
output_folder = "/mnt/scratch/ws/psbelokopytova/201907031108polinaB/3DPredictor/out/NPC/5KB/"
cell_type="NPC"
params = Parameters()
params.window_size = 5000 #region around contact to be binned for predictors
params.mindist = 10001 #minimum distance between contacting regions
params.maxdist = 1500000
params.sample_size = 1 #how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient=False
params.use_only_contacts_with_CTCF = "cont_with_CTCF"#"all_cont"#"#"all_cont"#"cont_with_CTCF "
write_all_chrms_in_file=True
fill_empty_contacts = False
logging.getLogger(__name__).debug("Using input folder "+input_folder)
from DataGenerator import generate_data
from PredictorGenerators import E1PredictorGenerator,ChipSeqPredictorGenerator, \
SmallChipSeqPredictorGenerator,SmallE1PredictorGenerator, SitesOrientPredictorGenerator, OrientBlocksPredictorGenerator, \
SitesOnlyOrientPredictorGenerator
start_time = time.time()
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "/home/evgeniy/asp/3Dpredictor/input/"
#output_folder = "D:/Users/Polina/3Dpredictor/"
output_folder = "/home/evgeniy/asp/3Dpredictor/out/"
#input_folder = "input"
params = Parameters()
params.window_size = 25000 #region around contact to be binned for predictors
#params.small_window_size = 12500 #region around contact ancors to be considered as cis
params.mindist = 50001 #minimum distance between contacting regions
#params.maxdist = params.window_size #max distance between contacting regions
params.maxdist = 1000000
#params.binsize = 20000 #when binning regions with predictors, use this binsize
params.sample_size = 500 #how many contacts write to file
params.conttype = "oe.gz"
training_file_name = "2018-09-23-trainingOrient.RandOnChr1." + str(
params) + ".txt"
validation_file_name = "validatingOrient." + str(params) + ".txt"
logging.getLogger(__name__).debug("Using input folder " + input_folder)
#Read contacts data
input_folder = args['input_folder']
output_folder = args['output_folder']
cell_type = args['cell_type']
start = int(args['start'])
end = int(args['end'])
chromosome = 'chr' + args['chr_num']
hic_name = args['hic_name']
CTCF_file_name = args['CTCF_file_name']
#RNA_file_name = args['RNA_file_name']
# validate_chrs = args['validate_chrs'].split(",")
# for chr in validate_chrs:
# chr = int(chr)
params = Parameters()
params.binsize = int(
args['binsize']
) #sequence resolution of contacts data. Use for finding of normalized coefficient file
params.window_size = params.binsize #region around contact to be binned for predictors. Usually equal to binsize
params.mindist = params.binsize * 2 + 1 #minimum distance between contacting regions
params.maxdist = 1500000
# params.sample_size = end - start
params.sample_size = 2 #how many contacts write to file
#params.conttype = conttype
params.max_cpus = int(args['max_cpus'])
params.keep_only_orient = False
params.use_only_contacts_with_CTCF = "all_cont" #"all_cont" or "cont_with_CTCF"
rearrangement = False
# deletion = Interval("chr" + chr_num, start, end)
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "/mnt/scratch/ws/psbelokopytova/202002281332polina_data_2019/3DPredictor/input/"
output_folder = "/mnt/scratch/ws/psbelokopytova/202002281332polina_data_2019/3DPredictor/out/mast_cells/"
cell_type = "mast_cells"
lengths_dict = {
'chr1': 1494930,
'chr3': 609806,
'chr5': 518646,
'chr7': 682860,
'chr11': 726290,
'chr13': 115324
}
params = Parameters()
params.binsize = 1000 #sequence resolution of contacts data. Use for finding of normalized coefficient file
params.window_size = params.binsize #region around contact to be binned for predictors. Usually equal to binsize
params.mindist = params.binsize * 2 + 1 #minimum distance between contacting regions
params.maxdist = 1500000
params.sample_size = 500000 #how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient = False
params.use_only_contacts_with_CTCF = "all_cont" #"all_cont"#"cont_with_CTCF"#"#"all_cont"#"cont_with_CTCF "
write_all_chrms_in_file = False #set True if you want write training file consisting several chromosomes
fill_empty_contacts = True #set True if you want use all contacts in region, without empty contacts
logging.getLogger(__name__).debug("Using input folder " + input_folder)
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "/mnt/scratch/ws/psbelokopytova/202001051010polina_data/3DPredictor/input/"
output_folder = "/mnt/scratch/ws/psbelokopytova/202001051010polina_data/3DPredictor/out/H1/"
cell_type = "K562"
lengths_dict = {
'chr1': 1494930,
'chr3': 609806,
'chr5': 518646,
'chr7': 682860,
'chr11': 726290,
'chr13': 115324
}
params = Parameters()
params.binsize = 1000 #sequence resolution of contacts data. Use for finding of normalized coefficient file
params.window_size = params.binsize #region around contact to be binned for predictors. Usually equal to binsize
params.mindist = params.binsize * 2 + 1 #minimum distance between contacting regions
params.maxdist = 1500000
params.sample_size = 500000 #how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient = False
params.use_only_contacts_with_CTCF = "all_cont" #"all_cont"#"cont_with_CTCF"#"#"all_cont"#"cont_with_CTCF "
write_all_chrms_in_file = False #set True if you want write training file consisting several chromosomes
fill_empty_contacts = False #set True if you want use all contacts in region, without empty contacts
logging.getLogger(__name__).debug("Using input folder " + input_folder)
1] #comma separated number of chromosomes for predictor generation
chr_nums = chr_num.split(",")
conttype = sys.argv[2] #contacts.gz or oe.gz
# chr_num="12,13,14"
# conttype = "contacts.gz"
if __name__ == '__main__': #Requiered for parallelization, at least on Windows
for conttype in [conttype]:
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "/input/K562/"
output_folder = "/output/K562/"
cell_type = "K562"
params = Parameters()
params.window_size = 5000 #region around contact to be binned for predictors
params.mindist = 10001 #minimum distance between contacting regions
params.maxdist = 1500000 #maximum distance between contacting regions
params.sample_size = 30000 #how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient = False #set True if you want use only CTCF with orient
params.use_only_contacts_with_CTCF = "all_cont" #"cont_with_CTCF" #this option use for training to change proportion
#of contacts with nearest ctcf sites
write_all_chrms_in_file = True #set True if you have train with few chromosomes. Need for writing different chromosomes in the same file
fill_empty_contacts = False
logging.getLogger(__name__).debug("Using input folder " + input_folder)
#Read contacts data
conttype = "contacts.gz" # contacts.gz or oe.gz
# chr_num="12,13,14"
# conttype = "contacts.gz"
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
if __name__ == '__main__': # Requiered for parallelization, at least on Windows
for conttype in [conttype]:
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = path1 + "/input/chr19_mm10/"
output_folder = path1 + "/output/chr19_mm10/"
cell_type = "NPC"
params = Parameters()
params.window_size = 25000 # region around contact to be binned for predictors
params.mindist = 50001 # minimum distance between contacting regions
params.maxdist = 1500000 # maximum distance between contacting regions
params.sample_size = 1 # how many contacts write to file
params.conttype = conttype
params.max_cpus = 11
params.keep_only_orient = False # set True if you want use only CTCF with orient
params.use_only_contacts_with_CTCF = "all_cont" # "cont_with_CTCF" #this option use for training to change proportion
# of contacts with nearest ctcf sites
write_all_chrms_in_file = False # set True if you have train with few chromosomes. Need for writing different chromosomes in the same file
fill_empty_contacts = False
logging.getLogger(__name__).debug("Using input folder " + input_folder)
# Read contacts data
import pandas as pd
import os
if __name__ == '__main__': #Requered for parallization, at least on Windows
#,"chr10", "chr1"]:
for conttype in ["contacts.gz", "oe.gz"]:
logging.basicConfig(format='%(asctime)s %(name)s: %(message)s',
datefmt='%I:%M:%S',
level=logging.DEBUG)
input_folder = "input/GM12878/"
#output_folder = "D:/Users/Polina/3Dpredictor/"
output_folder = "out/GM12878/"
#input_folder = "input"
params = Parameters()
params.window_size = 25000 #region around contact to be binned for predictors
#params.small_window_size = 12500 #region around contact ancors to be considered as cis
params.mindist = 50001 #minimum distance between contacting regions
#params.maxdist = params.window_size #max distance between contacting regions
params.maxdist = 1500000
#params.binsize = 20000 #when binning regions with predictors, use this binsize
params.sample_size = 250000 #how many contacts write to file
#params.conttype = "oe.gz"
params.conttype = conttype
params.max_cpus = 12
logging.getLogger(__name__).debug("Using input folder " + input_folder)
#Read contacts data
params.contacts_reader = ContactsReader()
