def get_qmatrix_data(qmatrix_path):
# check path exist
if utils.check_path_exist(qmatrix_path):
# initialize
qmatrix_comments_list = []
qmatrix_data_list = []
# open and read qmatrix
for qmatrix_line in open(qmatrix_path):
qmatrix_line = qmatrix_line.strip()
line_columns = qmatrix_line.split('\t')
# read data
if line_columns[0] == '*':
qmatrix_data_list.append(qmatrix_line)
else:
qmatrix_comments_list.append(qmatrix_line)
return (qmatrix_comments_list, qmatrix_data_list)
else:
sys.stderr.write("\n** Cannot open %s.\n" % qmatrix_path)
utils.die("** Program exit!")
def count_number_of_reads(reads_filepath, fasta_reads_flag):
# number of reads
number_of_reads = 0
# check path exist
if utils.check_path_exist(reads_filepath):
# check fasta or fastq format
if fasta_reads_flag: # fasta format
for read_line in open(reads_filepath,):
read_line = read_line.strip()
# fasta read id starts with >
if read_line.startswith('>'):
number_of_reads += 1
else: # fastq format
for read_line in open(reads_filepath):
read_line = read_line.strip()
# fastq read id starts with @
if read_line.startswith('@'):
number_of_reads += 1
else:
sys.stderr.write("\n** Cannot open %s.\n" % reads_filepath)
utils.die("** Program exit!")
return number_of_reads
def get_bootstrap_gvector(bootstrap_gvector_path):
# check path exist
if utils.check_path_exist(bootstrap_gvector_path):
# initialize
bootstrap_gvector_map = {}
# open and read bootstrap_gvector_path
for gvector_line in open(bootstrap_gvector_path):
gvector_line = gvector_line.strip()
gvector_columns = gvector_line.split('\t')
# read data
if gvector_columns[0] == '*':
genome_index = int(gvector_columns[1])
# ScaledPercentageChance is considered now.
# genome_chance = float(gvector_columns[2])
genome_chance = float(gvector_columns[3])
bootstrap_gvector_map[genome_index] = genome_chance
return bootstrap_gvector_map
else:
sys.stderr.write("\n** Cannot open %s!\n" % bootstrap_gvector_path)
utils.die("** Program exit!")
def check_aligned_reads(filename):
check_align = False
if utils.check_path_exist(filename):
check_align = True
return check_align
else:
sys.stderr.write("\n** Cannot open %s.\n" %(filename))
utils.die("** Program exit!")
def get_target_genome(gvector_path, reconstruction_selection, reconstruction_cutoff_abundance, reconstruction_genome_name):
# genome_name_map(key:genome_index, val:genome_name)
genome_name_map = {}
# genome_chance_map
genome_chance_map = {}
# target_genome_index_list
target_genome_index_list = []
# to check the reconstruction_genome_name
reconstruction_genome_name_exist_tag = False
# check g-vector file
if not utils.check_path_exist(gvector_path):
utils.die("** Error: No such file or directory: " + gvector_path)
# get target genome list (g_vector percentage >= cutoff_reconstruction_percentage)
with open(gvector_path, 'r') as f:
for gvector_line in f:
gvector_line = gvector_line.strip()
gvector_columns = gvector_line.split('\t')
# consider first column is '@': @ Genome_Index Genome_Name Alignment_Rate
if gvector_columns[0] == '@':
genome_index = int(gvector_columns[1])
genome_name = gvector_columns[2]
genome_name_map[genome_index] = genome_name
# consider first column is '*': * Genome_Index Percentage_Chance
elif gvector_columns[0] == '*':
genome_index = int(gvector_columns[1])
# ScaledPercentageChance is considered now.
# genome_chance = float(gvector_columns[2])
genome_chance = float(gvector_columns[3])
genome_chance_map[genome_index] = genome_chance
# reconstruct all genomes >= cut-off
if reconstruction_selection == 1:
# if genome_chance >= Reconstruction_Cutoff_Abundance, then save
if genome_chance >= reconstruction_cutoff_abundance:
target_genome_index_list.append(genome_index)
# reconstruct one target genome
else:
genome_name = genome_name_map[genome_index]
if genome_name == reconstruction_genome_name:
reconstruction_genome_name_exist_tag = True
target_genome_index_list.append(genome_index)
if not reconstruction_genome_name_exist_tag:
utils.die("** Check the config file! Reconstruction_Genome_Name value does not match to the SIGMA gvector results!")
if len(target_genome_index_list) == 0:
utils.die("** No target genomes exist for reconstruction!")
return (genome_name_map, genome_chance_map, target_genome_index_list)
def get_filtering_target_genome(gvector_path, filtering_genome_name):
# genome_name_map(key:genome_index, val:genome_name)
genome_name_map = {}
# genome_chance_map
genome_chance_map = {}
# target_genome_index_list
target_genome_index_list = []
# to check the filtering_genome_name
filtering_genome_name_exist_tag = False
# check g-vector file
if not utils.check_path_exist(gvector_path):
utils.die("** Error: No such file or directory: " + gvector_path)
# get target genome list
with open(gvector_path, 'r') as f:
for gvector_line in f:
gvector_line = gvector_line.strip()
gvector_columns = gvector_line.split('\t')
# consider first column is '@': @ Genome_Index Genome_Name Alignment_Rate
if gvector_columns[0] == '@':
genome_index = int(gvector_columns[1])
genome_name = gvector_columns[2]
genome_name_map[genome_index] = genome_name
# consider first column is '*': * Genome_Index Percentage_Chance
elif gvector_columns[0] == '*':
genome_index = int(gvector_columns[1])
# if PercentageChance is considered
genome_chance = float(gvector_columns[2])
# if ScaledPercentageChance is considered
#genome_chance = float(gvector_columns[3])
genome_chance_map[genome_index] = genome_chance
genome_name = genome_name_map[genome_index]
if genome_name == filtering_genome_name:
filtering_genome_name_exist_tag = True
target_genome_index_list.append(genome_index)
if not filtering_genome_name_exist_tag:
utils.die("** Check the config file! Filtering_Genome_Name value does not match to the SIGMA gvector results!")
if len(target_genome_index_list) == 0:
utils.die("** No target genomes exist for filtering!")
return (genome_name_map, genome_chance_map, target_genome_index_list)
def write_bootstrap_qmatrix(bootstrap_qmatrix_path, qmatrix_comments_list, qmatrix_data_list):
# check path exist
if utils.check_path_exist(bootstrap_qmatrix_path):
check_call(["rm", "-f", bootstrap_qmatrix_path], stdout = PIPE, stderr = sys.stderr)
bootstrap_qmatrix_out = open(bootstrap_qmatrix_path, 'wb')
# loop for qmatrix_comments_list
for qmatrix_comments_list_item in qmatrix_comments_list:
bootstrap_qmatrix_out.write(str(qmatrix_comments_list_item) + '\n')
# get qmatrix reads count
qmatrix_data_count = len(qmatrix_data_list)
# loop for qmatrix_data_list
for qmatrix_data_index in xrange(0, qmatrix_data_count):
random_index = random.randint(0, qmatrix_data_count - 1)
bootstrap_qmatrix_out.write(str(qmatrix_data_list[random_index]) + '\n')
def get_qmatrix_reads_count(qmatrix_path):
# check path exist
if utils.check_path_exist(qmatrix_path):
# initialize
qmatrix_reads_count = 0
# open and read qmatrix
for qmatrix_line in open(qmatrix_path):
qmatrix_line = qmatrix_line.strip()
line_columns = qmatrix_line.split('\t')
# read data
if line_columns[0] == '*':
qmatrix_reads_count += 1
return qmatrix_reads_count
else:
sys.stderr.write("\n** Cannot open %s.\n" % qmatrix_path)
utils.die("** Program exit!")
def get_gvector_genome_list(gvector_path):
# check file exist
if utils.check_path_exist(gvector_path):
# initialize
gvector_genome_list = []
# open file
gvector_file = open(gvector_path, 'r')
for gvector_line in gvector_file:
gvector_line = gvector_line.strip()
gvector_columns = gvector_line.split('\t')
# consider first column is '@': @ Genome_Index Genome_Name Alignment_Rate
if gvector_columns[0] == '@':
gvector_genome_list.append(gvector_line)
return gvector_genome_list
else:
sys.stderr.write("\n** Cannot open %s.\n" % qmatrix_path)
utils.die("** Program exit!")
def check_bowtie_index_built(genome_index_base,
genome_fasta_path_sublist):
# get bowtie index filepath (basename.1.bt2)
genome_index_bt1_path = genome_index_base + ".1.bt2"
# get fasta filepath (first fasta file)
genome_fasta_1_path = genome_fasta_path_sublist[0]
# check if the bowtie index file exist
if utils.check_path_exist(genome_index_bt1_path):
# get file creation time
index_creation_time = os.path.getctime(genome_index_bt1_path)
fasta_creation_time = os.path.getctime(genome_fasta_1_path)
# if index_creation_time is newer than fasta_creation_time
if index_creation_time >= fasta_creation_time:
return True
else:
return False
# if bowtie indexes do not exist, return false
else:
return False
def save(self, path=None, name='dqn_net.pkl'):
path = self.save_path if not path else path
utils.check_path_exist(path)
torch.save(self.eval_net.state_dict(), path + name)