def run(self):
"""Call ICE to cluster consensus isoforms."""
self.add_log("Start to run cluster.", level=logging.INFO)
#self.ice_opts.flnc_reads_per_split=1000 #FOR DEBUGGING, REMOVE LATER
# Split flnc_fa into smaller files and save files to _flnc_splitted_fas.
self.add_log("Splitting {flnc} into ".format(flnc=self.flnc_fa) +
"smaller files each containing {n} reads.".format(
n=self.ice_opts.flnc_reads_per_split),
level=logging.INFO)
self._flnc_splitted_fas = splitFasta(
input_fasta=self.flnc_fa,
reads_per_split=self.ice_opts.flnc_reads_per_split,
out_dir=self.root_dir,
out_prefix="input.split")
self.add_log("Splitted files are: " +
"\n".join(self._flnc_splitted_fas),
level=logging.INFO)
firstSplit = self._flnc_splitted_fas[0]
firstSplit_fq = firstSplit[:firstSplit.rfind('.')] + '.fastq'
self.add_log("Converting first split file {0} + {1} into fastq\n".format(\
firstSplit, self.ccs_fofn), level=logging.INFO)
# Convert this into FASTQ
ice_fa2fq(firstSplit, self.ccs_fofn, firstSplit_fq)
# Set up probabbility and quality value model
if self.ice_opts.use_finer_qv:
self._setProbQV_ccs(self.ccs_fofn, firstSplit)
else:
self._setProbQV_fq(firstSplitFq=firstSplit_fq)
# Initialize cluster by clique
self.add_log("Finding maximal cliques: initializing IceInit.",
level=logging.INFO)
self.iceinit = IceInit(readsFa=firstSplit,
qver_get_func=self._probqv.get_smoothed,
ice_opts=self.ice_opts,
sge_opts=self.sge_opts)
uc = self.iceinit.uc
# Dump uc to a file
self.add_log(
"Dumping initial clusters to {f}".format(f=self.initPickleFN),
level=logging.INFO)
with open(self.initPickleFN, 'w') as f:
cPickle.dump(uc, f)
# Run IceIterative.
self.add_log("Iterative clustering: initializing IceIterative.",
level=logging.INFO)
self.icec = IceIterative(
fasta_filename=firstSplit,
fasta_filenames_to_add=self._flnc_splitted_fas[1:],
all_fasta_filename=self.flnc_fa,
ccs_fofn=self.ccs_fofn,
root_dir=self.root_dir,
ice_opts=self.ice_opts,
sge_opts=self.sge_opts,
uc=uc,
probQV=self._probqv,
fastq_filename=firstSplit_fq,
use_ccs_qv=self.ice_opts.use_finer_qv)
self.add_log("IceIterative log: {f}.".format(f=self.icec.log_fn))
self.icec.run()
self.add_log("IceIterative completed.", level=logging.INFO)
# IceIterative done, write predicted (unplished) consensus isoforms
# to an output fasta
self.add_log("Creating a link to unpolished consensus isoforms.")
ln(self.icec.final_consensus_fa, self.out_fa)
# Call quiver to polish predicted consensus isoforms.
if self.ice_opts.quiver is not True:
self.add_log("Creating a link to cluster report.",
level=logging.INFO)
ln(src=self.icec.report_fn, dst=self.report_fn)
# Summarize cluster and write to summary_fn.
self.write_summary(summary_fn=self.summary_fn,
isoforms_fa=self.out_fa)
else: # self.ice_opts.quiver is True
self.add_log("Polishing clusters: initializing IcePolish.",
level=logging.INFO)
self.pol = Polish(root_dir=self.root_dir,
nfl_fa=self.nfl_fa,
bas_fofn=self.bas_fofn,
ccs_fofn=self.ccs_fofn,
fasta_fofn=self.fasta_fofn,
ice_opts=self.ice_opts,
sge_opts=self.sge_opts,
ipq_opts=self.ipq_opts,
nfl_reads_per_split=self.nfl_reads_per_split)
self.add_log("IcePolish log: {f}.".format(f=self.pol.log_fn),
level=logging.INFO)
self.pol.run()
self.add_log("IcePolish completed.", level=logging.INFO)
# cluster report
self.add_log("Creating a link to cluster report.",
level=logging.INFO)
ln(src=self.pol.iceq.report_fn, dst=self.report_fn)
# Summarize cluster & polish and write to summary_fn.
self.write_summary(summary_fn=self.summary_fn,
isoforms_fa=self.out_fa,
hq_fa=self.pol.icepq.quivered_good_fa,
lq_fa=self.pol.icepq.quivered_bad_fa)
# Create log file.
self.close_log()
return 0
def run(self):
"""Call ICE to cluster consensus isoforms."""
self.add_log("Start to run cluster.", level=logging.INFO)
# Split flnc_fa into smaller files and save files to _flnc_splitted_fas.
self.add_log("Splitting {flnc} into ".format(flnc=self.flnc_fa) +
"smaller files each containing {n} reads.".format(
n=self.ice_opts.flnc_reads_per_split),
level=logging.INFO)
self._flnc_splitted_fas = splitFasta(
input_fasta=self.flnc_fa,
reads_per_split=self.ice_opts.flnc_reads_per_split,
out_dir=self.root_dir,
out_prefix="input.split")
self.add_log("Splitted files are: " +
"\n".join(self._flnc_splitted_fas),
level=logging.INFO)
firstSplit = self._flnc_splitted_fas[0]
# Set up probabbility and quality value model
self._setProbQV(ccs_fofn=self.ccs_fofn, firstSplitFa=firstSplit)
# Initialize cluster by clique
# check if init.pickle already exists, if so, no need to run IceInit
if os.path.exists(self.initPickleFN):
self.add_log("{0} already exists. Reading to get uc.".format(
self.initPickleFN),
level=logging.INFO)
with open(self.initPickleFN) as f:
uc = cPickle.load(f)
else:
self.add_log("Finding maximal cliques.", level=logging.INFO)
self.iceinit = IceInit(readsFa=firstSplit,
qver_get_func=self._probqv.get_smoothed,
ice_opts=self.ice_opts,
sge_opts=self.sge_opts)
uc = self.iceinit.uc
# Dump uc to a file
self.add_log(
"Dumping initial clusters to {f}".format(f=self.initPickleFN),
level=logging.INFO)
with open(self.initPickleFN, 'w') as f:
cPickle.dump(uc, f)
# Run IceIterative.
self.add_log("Iteratively clustering.", level=logging.INFO)
self.icec = IceIterative(
fasta_filename=firstSplit,
fasta_filenames_to_add=self._flnc_splitted_fas[1:],
all_fasta_filename=self.flnc_fa,
ccs_fofn=self.ccs_fofn,
root_dir=self.root_dir,
ice_opts=self.ice_opts,
sge_opts=self.sge_opts,
uc=uc,
probQV=self._probqv)
self.icec.run()
clean_up_after_ICE(self.root_dir)
# IceIterative done, write predicted (unplished) consensus isoforms
# to an output fasta
self.add_log("Creating a link to unpolished consensus isoforms.")
ln(self.icec.final_consensus_fa, self.out_fa)
# Call quiver to polish predicted consensus isoforms.
if self.ice_opts.quiver is not True:
self.add_log("Creating a link to cluster report.")
ln(src=self.icec.report_fn, dst=self.report_fn)
self.add_log("Writing a summary to {f}".format(f=self.summary_fn),
level=logging.INFO)
self.writeSummary(fa=self.out_fa, summary_fn=self.summary_fn)
else: # self.ice_opts.quiver is True
#TODO review code
self.pol = Polish(root_dir=self.root_dir,
nfl_fa=self.nfl_fa,
bas_fofn=self.bas_fofn,
ccs_fofn=self.ccs_fofn,
hq_isoforms_fa=self.hq_isoforms_fa,
hq_isoforms_fq=self.hq_isoforms_fq,
lq_isoforms_fa=self.lq_isoforms_fa,
lq_isoforms_fq=self.lq_isoforms_fq,
ice_opts=self.ice_opts,
sge_opts=self.sge_opts)
self.pol.run()
# cluster report
self.add_log("Creating a link to cluster report.")
ln(src=self.pol.iceq.report_fn, dst=self.report_fn)
# Write a summary.
self.add_log("Writing a summary to {f}".format(f=self.summary_fn),
level=logging.INFO)
self.writeSummary(fa=self.out_fa,
summary_fn=self.summary_fn,
hq_fa=self.pol.icepq.quivered_good_fa,
lq_fa=self.pol.icepq.quivered_bad_fa)
# Create log file.
self.close_log()
return 0