def run(self):
"""
09-05-05
2006-09-21 add fuzzyDense_flag
2006-11-02 add tfbs_association_type
2006-11-02 differentiate good_cluster_table as pattern_xxx or good_xxx for pattern id
--db_connect()
--get_gene_no2bs_no_block()
--construct_two_dicts()
--input_node()
--fetch_cluster_block()
--computing_node()
--node_fire()
--cluster_bs_analysis()
--create_cluster_bs_table()
--output_node()
--submit_cluster_bs_table()
"""
communicator = MPI.world.duplicate()
node_rank = communicator.rank
free_computing_nodes = range(1,communicator.size-1)
print self.tfbs_association_type
if node_rank == 0:
(conn, curs) = db_connect(self.hostname, self.dbname, self.schema)
if self.tfbs_association_type==1: #2006-11-02
gene_no2bs_no_block = self.get_gene_no2bs_no_block(curs)
elif self.tfbs_association_type==2:
gene_no2bs_no_block = get_gene_no2bs_no_block_from_expt_tf_mapping(curs)
for node in range(1, communicator.size-1): #send it to the computing_node
communicator.send(gene_no2bs_no_block, node, 0)
if self.fuzzyDense_flag: #2006-09-21 add fuzzyDense_flag
#12-18-05 get edge2encodedOccurrence
MpiCrackSplat_instance = MpiCrackSplat()
edge2encodedOccurrence = {}
min_sup = 5 #need to expose them
max_sup = 40
total_vertex_set = self.return_total_vertex_set(curs, self.good_cluster_table)
edge2encodedOccurrence, no_of_datasets = self.fill_edge2encodedOccurrence(\
self.sig_vector_fname, min_sup, max_sup, total_vertex_set)
edge2encodedOccurrence_pickle = cPickle.dumps(edge2encodedOccurrence, -1)
for node in free_computing_nodes: #send it to the computing_node
communicator.send(edge2encodedOccurrence_pickle, node, 0)
elif node_rank>0 and node_rank<communicator.size-1:
data, source, tag, count = communicator.receive(Numeric.Int, 0, 0)
gene_no2bs_no_set, bs_no2gene_no_set = self.construct_two_dicts(node_rank, data)
if self.fuzzyDense_flag: #2006-09-21
#12-18-05
data, source, tag = communicator.receiveString(0, 0)
edge2encodedOccurrence = cPickle.loads(data)
elif node_rank==communicator.size-1: #establish connection before pursuing
(conn, curs) = db_connect(self.hostname, self.dbname, self.schema)
#12-20-05 for darwin output
gene_id2symbol = get_gene_id2gene_symbol(curs, self.tax_id)
dataset_no2desc = get_dataset_no2desc(curs)
mpi_synchronize(communicator)
if node_rank == 0:
if self.good_cluster_table.find('pattern')!=-1: #2006-11-02 it's pattern_xxx table, use id as pattern_id
curs.execute("DECLARE crs CURSOR FOR select distinct id, vertex_set, recurrence_array\
from %s "%(self.good_cluster_table))
else: #2006-11-02 it's good_xxx table, use mcl_id as pattern_id
curs.execute("DECLARE crs CURSOR FOR select distinct mcl_id, vertex_set, recurrence_array\
from %s "%(self.good_cluster_table))
input_node(communicator, curs, free_computing_nodes, self.size, self.report)
curs.execute("close crs")
elif node_rank<=communicator.size-2: #exclude the last node
if self.fuzzyDense_flag: #2006-09-21
fuzzyDense_instance = fuzzyDense(edge2encodedOccurrence)
else:
fuzzyDense_instance = None
parameter_list = [gene_no2bs_no_set, bs_no2gene_no_set, self.ratio_cutoff, \
self.top_number, self.p_value_cut_off, fuzzyDense_instance, self.degree_cut_off, self.fuzzyDense_flag]
computing_node(communicator, parameter_list, self.computing_node_handler, report=self.report)
elif node_rank==communicator.size-1:
#12-20-05 comment out
if self.new_table:
self.create_cluster_bs_table(curs, self.cluster_bs_table)
parameter_list = [curs, self.cluster_bs_table]
output_node(communicator, free_computing_nodes, parameter_list, self.submit_cluster_bs_table, report=self.report)
if self.commit:
curs.execute("end")
"""
recurrence_array = recurrence_array[1:-1].split(',')
recurrence_array = map(float, recurrence_array)
fuzzyDense_instance = fuzzyDense(edge2encodedOccurrence, debug)
core_vertex_ls, recurrent_and_on_datasets_ls = fuzzyDense_instance.get_core_vertex_set(vertex_list, recurrence_array, degree_cut_off)
from MpiClusterBsStat import MpiClusterBsStat
MpiClusterBsStat_instance = MpiClusterBsStat()
gene_no2bs_no_block = MpiClusterBsStat_instance.get_gene_no2bs_no_block(curs)
gene_no2bs_no_set, bs_no2gene_no_set = MpiClusterBsStat_instance.construct_two_dicts(0, gene_no2bs_no_block)
from TF_functions import cluster_bs_analysis
ls_to_return = cluster_bs_analysis(core_vertex_ls, gene_no2bs_no_set, bs_no2gene_no_set, ratio_cutoff, \
top_number, p_value_cut_off)
gene_id2symbol = get_gene_id2gene_symbol(curs, tax_id)
dataset_no2desc = get_dataset_no2desc(curs)
dataset_no_desc_ls = []
for dataset_index in recurrent_and_on_datasets_ls:
dataset_no = dataset_index +1
dataset_no_desc_ls.append([dataset_no, dataset_no2desc[dataset_no]])
outf = open(output_file, 'w')
outf.write("out:=[\n")
for i in range(len(ls_to_return)):
row = ls_to_return[i]
score, score_type, bs_no_list, target_gene_no_list, global_ratio, local_ratio, expected_ratio, unknown_ratio = row
core_vertex_symbol_ls = dict_map(gene_id2symbol, core_vertex_ls)
bs_no_symbol_list = dict_map(gene_id2symbol, bs_no_list)
if i == len(ls_to_return)-1:
