def process(n):
fasta_file = opts[FASTA]
fasta_file+="."+format(n-1,'05n')
file = os.path.abspath(fasta_file)
cmd = "ddb.pl -mode update -submode nr -taxid 1 -filename %s" % file
hddb.tunnel()
_debug(cmd)
subprocess.check_call(cmd,shell=True)
def process(n):
fasta_file = opts[FASTA]
fasta_file+=".%i" % n-1
file = os.path.abspath(fasta_file)
cmd = "ddb.pl -mode update -submode nr -filename %s" % file
hddb.tunnel()
_debug(cmd)
subprocess.check_call(cmd,shell=True)
cmd = "ddb.pl -mode update -submode nr"
def protein(sequence_key):
#subprocess.check_call("nohup `which Xvfb` :20 -screen 0 1024x768x24 &> /dev/null & sleep 7",shell=True)
print >>sys.stderr, os.getenv("HOSTNAME")+" "+os.getenv("SGE_TASK_ID")
hddb.tunnel()
#-Djava.awt.headless=true
#DISPLAY=localhost:20.0
cmd = "ddb.pl -mode export -submode ginzu2yeastrc -ginzu 1 -parent_sequence_key %i" % sequence_key
subprocess.check_call(cmd,shell=True)
return sequence_key
def remcm(sequence_key):
import sys
import os
from hpf.hddb import tunnel, kill_tunnel
tunnel(5)
from hpf.mcm.process import McmReRun
from hpf.mcm import McmDB
with open("/scratch/kd35/db175/data.mammothDb") as handle:
mcmdb = McmDB.load(handle)
mcmdb.scop='1.75'
mcmdb.list_file = "/scratch/kd35/db175/list.mammoth"
# Add int keys to the mcmdb._dict (this is rough)
for key in mcmdb._dict.keys():
mcmdb._dict[int(key)] = mcmdb._dict[key]
jobdir = "/scratch/kd35/tmp/mcm/jobs/%i" % int(sequence_key)
try:
# Create an McmReRun task and run it
if not os.path.exists(jobdir):
os.mkdir(jobdir)
os.chdir(jobdir)
log_file = "/scratch/kd35/tmp/mcm/jobs/%i/outlog" % int(sequence_key)
out_handle = open(log_file,"w")
runtime().set_log(stdout=out_handle, stderr=out_handle)
logging.basicConfig(filename=log_file, filemode='a')
runtime().debug(os.environ['HOSTNAME'])
task = McmReRun(jobdir,mcmdb,int(sequence_key))
task.run()
finally:
# Clean up: gzip results, save log file, remove all other tmp files
import subprocess
runtime().debug("finally")
for file in os.listdir(jobdir):
if "data.mammoth" in file:
#save and gzip
subprocess.Popen([r"gzip","-f",file])
elif "outlog" in file:
runtime().debug("saving outlog")
else:
os.remove(os.path.join(jobdir,file))
#os.rmdir(jobdir)
kill_tunnel()
def main(*args):
pool = processor(synchronous=runtime().opt(SYNCHRONOUS))
runtime().debug("Using processor",pool)
pool.make_tasks(None)
consume(pool.run(remcm))
def _do(argv):
r = runtime()
r.description("""
remcm.py [-options]
Re-MCM a set of sequence keys. Expects a 'tasks.pickle'.
""")
r.add_option(Flag(SYNCHRONOUS, "s", description="Run this script synchronously without any multi-processing"))
args = r.parse_options(argv)
main(*args)
if __name__=="__main__":
_do(sys.argv[1:])
#!/usr/bin/env python
'''
Created on Apr 6, 2010
Takes a list of prediction codes to set up input files and create a task pickle
for running MCM.
Originally used to re-do the since solved.
@author: Patrick
'''
from hpf.hddb import tunnel
tunnel(sleep=1)
import sys, os
import getopt
import shutil
from hpf.runtime import runtime, Flag, FileOption, debug, DEBUG
from hpf.utilities import consume
from hpf.processing import processor, SYNCHRONOUS
from hpf.mcm.mcm import SCRATCH
from hpf.hddb.db import *
FRAGMENTS = "fragments"
SCRATCH = "SCRATCH"
HPF1_DECOYS = "hpf1_decoys"
HPF2_DECOYS = "hpf2_decoys"
def prep(prediction_code):
runtime().debug("Running on",prediction_code)
with PrepPredictionCode(prediction_code,
runtime().opt(SCRATCH),
runtime().opt(HPF1_DECOYS),
def _connection(db="hpf"):
global _db
if _db == None:
tunnel(5)
_db = MySQLdb.connect(host="127.0.0.1",port=13307,db=db,user="******",passwd="bonneaulab")
return _db