def get_basic_cluster_dstructure(self, curs, mcl_id, splat_table, mcl_table):
"""
04-06-05
"""
sys.stderr.write("Getting the basic information of cluster no.%s..."%mcl_id)
unit = cluster_dstructure()
curs.execute("select m.mcl_id, m.vertex_set, m.connectivity, 0,\
m.recurrence_array, s.edge_set, s.connectivity from %s m, %s s where m.splat_id=s.splat_id and \
m.mcl_id=%s"\
%(mcl_table, splat_table, mcl_id)) #06-20-05 connectivity_original faked to be 0
rows = curs.fetchall()
if len(rows)>0:
for row in rows:
unit.cluster_id = row[0]
vertex_set = row[1][1:-1].split(',')
unit.vertex_set = map(int, vertex_set)
unit.connectivity = row[2]
unit.connectivity_original = row[3]
recurrence_array = row[4][1:-1].split(',')
unit.recurrence_array = map(float, recurrence_array)
unit.edge_set = parse_splat_table_edge_set(row[5])
unit.splat_connectivity = row[6]
sys.stderr.write("Done.\n")
else:
unit = None
sys.stderr.write("Cluster: %s not found.\n"%mcl_id)
return unit
def copath_parser(self, row, writer, argument=None, argument2=None):
"""
04-12-05
copied from codense2db.py, changed a lot
"""
cooccurrent_cluster_id = self.p_cooccurrent_cluster_id.match(row[0]).group()
vertex_set = row[2][1:-2].split(";")
vertex_set = map(int, vertex_set)
edge_list = row[3][2:-4].split(" );(")
edge_set = []
for edge in edge_list:
edge = edge.split(",")
edge = map(int, edge)
# in ascending order
edge.sort()
edge_set.append(edge)
# 04-29-05 cc module come into play to get the connected components
instance = cc_from_edge_list()
instance.run(edge_set)
cc_list = instance.cc_list
for cc_edge_list in cc_list:
cluster = cluster_dstructure()
cluster.cooccurrent_cluster_id = cooccurrent_cluster_id # it's not used in the output()
# initialize two sets
cluster.vertex_set = self.vertex_set_from_cc_edge_list(cc_edge_list)
cluster.edge_set = cc_edge_list
self.output(writer, cluster)
def data_fetch(self, curs, splat_table, mcl_table, crs_no=0, output_fname=None):
"""
04-17-05
fetch cluster_dstructures for all clusters(Jasmine's request)
04-19-05
1. return a mcl_id2cluster_dstructure
2. crs_no
08-31-05
output clusters directly to output_fname
09-01-05
add the last []
"""
gene_no2gene_id = get_gene_no2gene_id(curs) #08-31-05
outf = open(output_fname, 'w') #08-31-05
outf.write("r:=[") #08-31-05
mcl_id2cluster_dstructure = {}
no_of_total_genes = get_no_of_total_genes(curs)
sys.stderr.write("Getting the basic information for all clusters...\n")
curs.execute("DECLARE crs%s CURSOR FOR select m.mcl_id, m.vertex_set, m.connectivity, 0,\
m.recurrence_array, s.edge_set, s.connectivity, m.cooccurrent_cluster_id from %s m, %s s where \
m.splat_id=s.splat_id"\
%(crs_no, mcl_table, splat_table)) #06-20-05 connectivity_original faked to be 0
curs.execute("fetch 5000 from crs%s"%crs_no)
rows = curs.fetchall()
while rows:
for row in rows:
unit = cluster_dstructure()
unit.cluster_id = row[0]
vertex_set = row[1][1:-1].split(',')
unit.vertex_set = map(int, vertex_set)
unit.connectivity = row[2]
unit.connectivity_original = row[3]
recurrence_array = row[4][1:-1].split(',')
unit.recurrence_array = map(float, recurrence_array)
unit.edge_set = parse_splat_table_edge_set(row[5])
unit.splat_connectivity = row[6]
unit.cooccurrent_cluster_id = row[7]
unit.go_no2association_genes = self.get_go_functions_of_this_gene_set(curs, unit.vertex_set)
unit.go_no2information = self.get_information_of_go_functions(curs, \
unit.go_no2association_genes, len(unit.vertex_set), no_of_total_genes, p_value_cut_off=0.05) #jasmine wants to cut some go-nos.
unit.edge_cor_2d_list, unit.edge_sig_2d_list = self.get_cor_sig_2d_list(curs, unit.edge_set)
str_tmp = self.return_string_form_of_cluster_dstructure(unit, gene_no2gene_id) #08-31-05
outf.write("%s,"%str_tmp)
#mcl_id2cluster_dstructure[unit.cluster_id] = unit
"""
order_1st_id, order_2nd_id = map(int, unit.cooccurrent_cluster_id.split('.'))
if order_1st_id not in self.order_1st_id2all_clusters:
self.order_1st_id2all_clusters[order_1st_id] = {}
if order_2nd_id not in self.order_1st_id2all_clusters[order_1st_id]:
self.order_1st_id2all_clusters[order_1st_id][order_2nd_id] = []
self.order_1st_id2all_clusters[order_1st_id][order_2nd_id].append(unit)
"""
curs.execute("fetch 5000 from crs%s"%crs_no)
rows = curs.fetchall()
outf.write("[]]:") #08-31-05, 09-01-05 add the last blank []
del outf
sys.stderr.write("Done.\n")
return mcl_id2cluster_dstructure
def copath_parser(self, row, argument=None, argument2=None):
"""
04-12-05
copied from codense2db.py, changed a lot
"""
cooccurrent_cluster_id = self.p_cooccurrent_cluster_id.match(row[0]).group()
vertex_set = row[2][1:-2].split(';')
vertex_set = map(int, vertex_set)
edge_list = row[3][2:-4].split(' );(')
edge_set = []
for edge in edge_list:
edge = edge.split(',')
edge = map(int, edge)
#in ascending order
edge.sort()
edge_set.append(edge)
if cooccurrent_cluster_id not in self.cooccurrent_cluster_id2cluster:
cluster = cluster_dstructure()
cluster.cooccurrent_cluster_id = cooccurrent_cluster_id #it's not used in the output()
#initialize two sets
cluster.vertex_set = []
cluster.edge_set = []
self.cooccurrent_cluster_id2cluster[cooccurrent_cluster_id] = cluster
#pass it to ease programming
cluster = self.cooccurrent_cluster_id2cluster[cooccurrent_cluster_id]
cluster.vertex_set += vertex_set
cluster.edge_set += edge_set
