def run(blast_filename):
filename = c.step1_process_blast('',blast_filename,30)
return
g = read_gpickle(filename)
g.delete_nodes_from(filter(lambda n: g.degree(n)<=2, g.nodes_iter()))
g_nodes = g.nodes()
print >> sys.stderr, "number of nodes...",g.number_of_nodes()
S, H = p.convert_graph_connectivity_to_sparse(g, g_nodes)
valid_rows = range(g.number_of_nodes())
QQ = []
last_QQ_size = 0
failed_count = 0
while failed_count < 10:
p.get_cliques(S, H, QQ, 0.8, 10, valid_rows)
if len(QQ) == last_QQ_size:
failed_count += 1
else:
failed_count = 0
last_QQ_size += 1
f = open(blast_filename+'.gamma80iter10.cliques','w')
for q in QQ:
f.write(" ".join(map(lambda x: g_nodes[x], q))+"\n")
print >> sys.stderr, "output written to....",f.name
f.close()
import os,re,sys
from networkx import *
from networkx.readwrite import *
from cPickle import *
import cluster_steps as c
from miscMySQL import *
dirname = 'output/output_blast_etc/ALLFirm_RefSeq25_m30s0/'
conn = get_conn_Firm()
cursor = conn.cursor()
nodes_to_index = c.get_nodes_to_index_from_db(cursor)
G = Graph()
while G.number_of_edges() < 2100000:
cursor.execute("select filename from files_todo order by rand() limit 1")
try:
filename = cursor.fetchone()[0]
except:
print >> sys.stderr, "couldn't fetch a file to process....no more?"
break
cursor.execute("delete from files_todo where filename=\"{0}\"".format(filename))
print >> sys.stderr, filename
c.step1_process_blast(dirname+filename, 35, nodes_to_index, G)
