def compare_gene_condition_vertex_set(curs, p_gene_table, gene_p_table, good_cluster_table, output_fname):
import os, sys, csv
from sets import Set
from codense.common import pg_1d_array2python_ls
sys.stderr.write("Getting gene_no2mcl_id_go_no_list ...\n")
gene_no2mcl_id_go_no_list = {}
curs.execute(
"DECLARE crs1 CURSOR for select p.gene_no, p.mcl_id, p.go_no from %s p, %s g\
where p.p_gene_id=g.p_gene_id"
% (p_gene_table, gene_p_table)
)
counter = 0
curs.execute("fetch 1000 from crs1")
rows = curs.fetchall()
mcl_id_set = Set()
while rows:
for row in rows:
gene_no, mcl_id, go_no = row
mcl_id_set.add(mcl_id)
if gene_no not in gene_no2mcl_id_go_no_list:
gene_no2mcl_id_go_no_list[gene_no] = []
gene_no2mcl_id_go_no_list[gene_no].append([mcl_id, go_no])
counter += 1
sys.stderr.write("%s%s" % ("\x08" * 30, counter))
curs.execute("fetch 1000 from crs1")
rows = curs.fetchall()
curs.execute("close crs1")
sys.stderr.write("Done.\n")
sys.stderr.write("Getting mcl_id2recurrence_array_vertex_set ...\n")
mcl_id2recurrence_array_vertex_set = {}
curs.execute("DECLARE crs0 CURSOR for select mcl_id, vertex_set, recurrence_array from %s" % good_cluster_table)
counter = 0
curs.execute("fetch 1000 from crs0")
rows = curs.fetchall()
while rows:
for row in rows:
mcl_id, vertex_set, recurrence_array = row
vertex_set = pg_1d_array2python_ls(vertex_set)
recurrence_array = pg_1d_array2python_ls(recurrence_array)
mcl_id2recurrence_array_vertex_set[mcl_id] = [recurrence_array, Set(vertex_set)]
counter += 1
sys.stderr.write("%s%s" % ("\x08" * 30, counter))
curs.execute("fetch 1000 from crs0")
rows = curs.fetchall()
curs.execute("close crs0")
sys.stderr.write("Done.\n")
sys.stderr.write("Comparing gene condition and vertex_set ...\n")
writer = csv.writer(open(output_fname, "w"), delimiter="\t")
for gene_no, mcl_id_go_no_list in gene_no2mcl_id_go_no_list.iteritems():
cmp_condition_vertex_result = get_mcl_id_sharing(mcl_id_go_no_list, mcl_id2recurrence_array_vertex_set)
for row in cmp_condition_vertex_result:
writer.writerow([gene_no] + row)
sys.stderr.write("Done.\n")
def get_prot_interaction_graph(self, curs, prot_interaction_table, tax_id):
"""
2006-11-20
2006-01-05
print the number of connected components in the end
"""
sys.stderr.write("Getting protein interaction graph ...\n")
curs.execute("DECLARE crs0 CURSOR for select gene_id_array, interaction_type_id\
from %s where tax_id=%s"%(prot_interaction_table, tax_id))
curs.execute("fetch 3000 from crs0")
rows = curs.fetchall()
prot_interaction_graph = nx.XGraph()
counter = 0
while rows:
for row in rows:
gene_id_array, interaction_type_id = row
gene_id_array = pg_1d_array2python_ls(gene_id_array)
if len(gene_id_array)>1:
for i in range(len(gene_id_array)):
for j in range(i+1, len(gene_id_array)):
prot_interaction_graph.add_edge(gene_id_array[i], gene_id_array[j], interaction_type_id)
counter += 1
if self.report:
sys.stderr.write("%s%s"%('\x08'*20, counter))
curs.execute("fetch 3000 from crs0")
rows = curs.fetchall()
curs.execute("close crs0")
sys.stderr.write("%s nodes, %s edges and %s components, Done.\n"%(nx.number_of_nodes(prot_interaction_graph),\
nx.number_of_edges(prot_interaction_graph), nx.number_connected_components(prot_interaction_graph)))
return prot_interaction_graph
def get_prom_seq_from_entrezgene_mapping_table(self, curs, prom_seq_table, entrezgene_mapping_table='entrezgene_mapping', \
annot_assembly_table = 'annot_assembly'):
sys.stderr.write("Getting prom_seq from entrezgene_mapping_table...\n")
curs.execute("DECLARE crs CURSOR FOR SELECT e.gene_id, e.genomic_gi, e.tax_id, a.chromosome, e.strand,\
e.start, e.stop, e.mrna_start, e.mrna_stop, e.cds_start, e.cds_stop from %s e, %s a \
where e.genomic_gi=a.gi"%(entrezgene_mapping_table, annot_assembly_table))
curs.execute("fetch 10000 from crs")
rows = curs.fetchall()
counter = 0
while rows:
for row in rows:
gene_id, genomic_gi, tax_id, chromosome, strand, start, stop, mrna_start, mrna_stop, cds_start, cds_stop = row
seg_loc_ls = []
if cds_start and cds_stop:
cds_start = pg_1d_array2python_ls(cds_start, int)
cds_stop = pg_1d_array2python_ls(cds_stop, int)
for i in range(len(cds_start)):
seg_loc_ls.append([cds_start[i],cds_stop[i]])
elif mrna_start and mrna_stop:
mrna_start = pg_1d_array2python_ls(mrna_start, int)
mrna_stop = pg_1d_array2python_ls(mrna_stop, int)
for i in range(len(mrna_start)):
seg_loc_ls.append([mrna_start[i],mrna_stop[i]])
else:
seg_loc_ls.append([start, stop])
seg_loc_ls.sort() #some genes have reversed cds order
ps_attr_instance = prom_seq_attr()
ps_attr_instance.prom_acc = gene_id
ps_attr_instance.chromosome = chromosome
ps_attr_instance.organism = tax_id2org(tax_id)
upstream_loc_ls = [0,0]
instron_1st_loc_ls = []
if strand=='1': #plus strand
ps_attr_instance.strand = '+'
upstream_loc_ls[1] = seg_loc_ls[0][0]-1
upstream_loc_ls[0] = upstream_loc_ls[1] - 9999
if upstream_loc_ls[0]<1: #in case exceed the chromosome boundary
upstream_loc_ls[0] = 1
#check whether there's gene upstream
upstream_loc_ls[0] = self.return_closest_anchor(curs, 'stop', upstream_loc_ls, gene_id, tax_id, genomic_gi, \
entrezgene_mapping_table)
if upstream_loc_ls[0]>upstream_loc_ls[1]: #No upstream
if self.debug:
sys.stderr.write("\tgene_id: %s no upstream\n"%gene_id)
upstream_loc_ls = []
if len(seg_loc_ls)>1: #the first intron
instron_1st_loc_ls.append(seg_loc_ls[0][1]+1)
instron_1st_loc_ls.append(seg_loc_ls[1][0]-1)
elif strand=='-1': #minus strand
ps_attr_instance.strand = '-'
upstream_loc_ls[0] = seg_loc_ls[-1][1]+1
upstream_loc_ls[1] = upstream_loc_ls[0] + 9999
#NOTE: exceeding the chromosome boundary is taken care of by get_sequence_segment()
#check whether there's gene upstream
upstream_loc_ls[1] = self.return_closest_anchor(curs, 'start', upstream_loc_ls, gene_id, tax_id, genomic_gi, \
entrezgene_mapping_table)
if upstream_loc_ls[0]>upstream_loc_ls[1]: #No upstream
sys.stderr.write("\tgene_id: %s no upstream\n"%gene_id)
upstream_loc_ls = []
if len(seg_loc_ls)>1: #the first intron
instron_1st_loc_ls.append(seg_loc_ls[-2][1]+1)
instron_1st_loc_ls.append(seg_loc_ls[-1][0]-1)
else: #ignore genes with no strand info, some are not real genes
continue
#1st deal with upstream_loc_ls
if upstream_loc_ls:
ps_attr_instance.prom_genome_start = upstream_loc_ls[0]
ps_attr_instance.prom_genome_end = upstream_loc_ls[1]
ps_attr_instance.prom_type_id = 1
ps_attr_instance.sequence = get_sequence_segment(curs, genomic_gi, upstream_loc_ls[0], upstream_loc_ls[1])
self.submit_to_prom_seq(curs, prom_seq_table, ps_attr_instance)
#2nd handle instron_1st_loc_ls, might not exist
if instron_1st_loc_ls:
if instron_1st_loc_ls[0]>instron_1st_loc_ls[1]:
sys.stderr.write("\tgene_id: %s weird 1st intron %s.\n"%(gene_id, instron_1st_loc_ls))
ps_attr_instance.prom_genome_start = instron_1st_loc_ls[0]
ps_attr_instance.prom_genome_end = instron_1st_loc_ls[1]
ps_attr_instance.prom_type_id = 5
ps_attr_instance.sequence = get_sequence_segment(curs, genomic_gi, instron_1st_loc_ls[0], instron_1st_loc_ls[1])
self.submit_to_prom_seq(curs, prom_seq_table, ps_attr_instance)
counter += 1
if self.report:
sys.stderr.write("%s\t%s"%('\x08'*20, counter))
if self.debug: #enough
break
curs.execute("fetch 10000 from crs")
rows = curs.fetchall()
sys.stderr.write("Done getting prom_seq from entrezgene_mapping_table.\n")
