def submit_qsub(qsub_opt, qsub_cmd, qsub_name, output_path, log):
job_id = 0
job_share = 150
cluster_project = "gentech-rqc.p"
qsub_path = os.path.join(output_path, "qsub")
if not os.path.isdir(qsub_path):
os.makedirs(qsub_path)
output_log = os.path.join(qsub_path, "qsub-%s.log" % qsub_name)
my_qsub = "qsub -b y -j y -m n -w e -terse -N %s -P %s -o %s -js %s %s '%s'" % (
qsub_name, cluster_project, output_log, job_share, qsub_opt, qsub_cmd)
# append to qsub.txt
cmd = "module load uge;%s" % my_qsub
stdOut, stdErr, exitCode = run_sh_command(cmd, True, log)
post_mortem_cmd(cmd, exitCode, stdOut, stdErr, log)
if exitCode == 0:
job_id = int(stdOut.strip())
log.info("- cluster job id: %s", job_id)
if job_id > 0:
qsub_log = os.path.join(output_path, "qsub_list.txt")
fh = open(qsub_log, "a")
fh.write("%s,%s\n" % (job_id, "submitted"))
fh.close()
return job_id
def read_qc(odir, fastq, status):
log_and_print(
"\n\n%s - RUN QC <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<%s\n"
% (color['pink'], color['']))
if not os.path.isfile(fastq):
return False, status
qcdir = os.path.join(odir, READQC_DIR)
if STEP_ORDER[status] < STEP_ORDER[QC_END]:
checkpoint(QC_START, status)
qcexe = os.path.join(os.path.dirname(__file__), 'readqc.py')
cmd = '%s -o %s -f %s --skip-blast' % (qcexe, qcdir, fastq)
# print('DEBUG : %s' % cmd)
stdOut, stdErr, exitCode = run_sh_command(cmd, True)
if exitCode != 0:
print('ERROR : %s' % stdErr)
return False, qcdir, status
checkpoint(QC_END, status)
status = QC_END
else:
log_and_print("No need to do qc step.")
return True, qcdir, status
def localize_file(fileNameFullPath, log):
done = 0
loopCount = 0
while done == 0:
loopCount += 1
if os.path.isfile(fileNameFullPath):
done = 1 # congratulations! The file system seems to be working for the moment
else:
sleepTime = 60 * loopCount
log.error(
"Filename doesn't exist on the system: %s, sleeping for %s seconds.",
fileNameFullPath, sleepTime)
time.sleep(sleepTime)
if loopCount > 3:
done = 1 # probably failed
if not os.path.isfile(fileNameFullPath):
return None
destinationPath = "/scratch/rqc/localized-file"
if not os.path.isdir(destinationPath):
# make_dir_p(destinationPath)
_, _, exitCode = run_sh_command(
"mkdir -p /scratch/rqc/localized-file && chmod 777 /scratch/rqc/localized-file",
True, log, True)
assert exitCode == 0
fileName = safe_basename(fileNameFullPath, log)[0]
localizeCmd = "rsync -av --omit-dir-times --no-perms " + fileNameFullPath + " " + os.path.join(
destinationPath, fileName)
_, _, exitCode = run_sh_command(localizeCmd, True, log, True)
localizedFileNameFullPath = None
if exitCode == RQCExitCodes.JGI_SUCCESS:
localizedFileNameFullPath = os.path.join(destinationPath, fileName)
log.info("File localization completed for %s" % (fileName))
else:
localizedFileNameFullPath = None
log.error("File localization failed for %s" % (fileName))
return localizedFileNameFullPath
def watch_cluster(output_path, log, sleep_time=300):
log.info("watch_cluster")
is_job_complete = "/usr/common/usg/bin/isjobcomplete.new" # + job_id = nice way to ask if job is done
qsub_log = os.path.join(output_path, "qsub_list.txt")
#sleep_time = 300 # check isjobrunning every xx seconds
hb_max = (
180 * 3600
) / sleep_time # total number of heartbeats before we give up, 180 hours worth of run time
#hb_max = 5 # temp
done = 0
hb_cnt = 0 # heartbeat count
if not os.path.isfile(qsub_log):
done = 1
while done == 0:
hb_cnt += 1
log.info("- heartbeat: %s", hb_cnt)
qsub_list = []
qsub_cnt = 0
qsub_complete = 0
qsub_err = 0
fh = open(qsub_log, "r")
for line in fh:
qsub_list.append(line.strip())
qsub_cnt += 1
fh.close()
fh = open(qsub_log, "w")
for qsub in qsub_list:
job_id, status = qsub.split(",")
new_status = status
if status in ("complete", "fail"):
continue
else:
cmd = "%s %s" % (is_job_complete, job_id)
stdOut, stdErr, exitCode = run_sh_command(cmd, True, log)
#post_mortem_cmd(cmd, exitCode, stdOut, stdErr, log)
running = "%s queued/running" % job_id
not_running = "%s not queued/running" % job_id
error_qw = "%s queued/running/error" % job_id
if stdOut.strip == running:
new_status = "running"
elif stdOut.strip() == not_running:
new_status = "complete" # might have failed
qsub_complete += 1
elif stdOut.strip() == error_qw:
new_status = "error"
qsub_err += 1
if stdOut.strip() == "not queued/running":
new_status = "complete"
fh.write("%s,%s\n" % (job_id, new_status))
log.info("- job_id: %s, status: %s", job_id, new_status)
fh.close()
qsub_running = qsub_cnt - (qsub_complete + qsub_err)
log.info("- job count: %s, running: %s, err: %s", qsub_cnt,
qsub_running, qsub_err)
if qsub_cnt == (qsub_complete + qsub_err):
done = 1
if hb_cnt >= hb_max:
done = 1
if done == 0:
time.sleep(sleep_time)
def localize_dir2(db, log, bbuffer=None):
safeBaseName, exitCode = safe_basename(db, log)
safeDirName, exitCode = safe_dirname(db, log)
targetScratchDir = None
nerscDb = "/scratch/blastdb/global/dna/shared/rqc/ref_databases/ncbi/CURRENT"
# nerscDb = "/" ## temporarily do not use NERSC db
## nerscDbPersistentBBuffer = "/var/opt/cray/dws/mounts/batch/NCBI_DB_striped_scratch/ncbi"
## check if persistent burst buffer is ready
if bbuffer is None and "DW_PERSISTENT_STRIPED_NCBI_DB" in os.environ and os.environ[
'DW_PERSISTENT_STRIPED_NCBI_DB'] is not None:
nerscDb = os.path.join(os.environ['DW_PERSISTENT_STRIPED_NCBI_DB'],
"ncbi")
else:
targetScratchDir = "/scratch/rqc"
if bbuffer is not None:
if not os.path.isdir(bbuffer):
log.error("Burst Buffer does not initiated: %s", bbuffer)
return None, RQCExitCodes.JGI_FAILURE
else:
targetScratchDir = bbuffer
log.info("Localization will use Burst Buffer location at %s",
targetScratchDir)
elif not os.path.isdir(targetScratchDir):
_, _, exitCode = run_sh_command(
"mkdir %s && chmod 777 %s" %
(targetScratchDir, targetScratchDir), True, log, True)
assert exitCode == 0
rsyncOption = ""
src = db
dest = ""
## For
# GREEN_GENES = "/global/dna/shared/rqc/ref_databases/misc/CURRENT/green_genes16s.insa_gg16S.fasta"
# LSU_REF = "/global/dna/shared/rqc/ref_databases/misc/CURRENT/LSURef_115_tax_silva.fasta"
# SSU_REF = "/global/dna/shared/rqc/ref_databases/misc/CURRENT/SSURef_NR99_115_tax_silva.fasta"
# LSSU_REF = "/global/dna/shared/rqc/ref_databases/misc/CURRENT/LSSURef_115_tax_silva.fasta"
# CONTAMINANTS = "/global/dna/shared/rqc/ref_databases/misc/CURRENT/JGIContaminants.fa"
# COLLAB16S = "/global/dna/shared/rqc/ref_databases/misc/CURRENT/collab16s.fa"
if db.endswith(".fa") or db.endswith(".fasta"):
rsyncOption = "--include '*.n??' --exclude '*.fa' --exclude '*.fasta' --exclude '%s' --exclude '*.log'" % (
safeBaseName)
src = db + ".n??"
dest = os.path.join(targetScratchDir, safeBaseName)
blastDb = dest
if os.path.isfile(
dest): ## just in case for the file name already exists
rmCmd = "rm -rf %s" % (dest)
run_sh_command(rmCmd, True, log, True)
## For
# NT_maskedYindexedN_BB = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/nt/bbtools_dedupe_mask/nt_bbdedupe_bbmasked_formatted"
elif db.endswith("nt_bbdedupe_bbmasked_formatted"):
if os.path.isdir(os.path.join(nerscDb, "nt/bbtools_dedupe_mask")):
blastDb = os.path.join(nerscDb, "nt/bbtools_dedupe_mask")
log.info("NERSC NCBI Database found: %s" % (blastDb))
return blastDb, RQCExitCodes.JGI_SUCCESS
rsyncOption = "--include '*.n??' --exclude '*.fna' --exclude 'nt_bbdedupe_bbmasked_formatted' --exclude '*.log'"
src = safeDirName + '/'
dest = os.path.join(targetScratchDir, safeBaseName)
blastDb = dest
## For
# NR = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/nr/nr"
# REFSEQ_ARCHAEA = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.archaea/refseq.archaea"
# REFSEQ_BACTERIA = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.bacteria/refseq.bacteria"
# REFSEQ_FUNGI = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.fungi/refseq.fungi"
# REFSEQ_MITOCHONDRION = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.mitochondrion/refseq.mitochondrion"
# REFSEQ_PLANT = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.plant/refseq.plant"
# REFSEQ_PLASMID = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.plasmid/refseq.plasmid"
# REFSEQ_PLASTID = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.plastid/refseq.plastid"
# REFSEQ_VIRAL = "/global/dna/shared/rqc/ref_databases/ncbi/CURRENT/refseq.viral/refseq.viral"
else:
if os.path.isdir(os.path.join(nerscDb, safeBaseName)):
blastDb = os.path.join(nerscDb, safeBaseName)
log.info("NERSC NCBI Database found: %s" % (blastDb))
return blastDb, RQCExitCodes.JGI_SUCCESS
rsyncOption = "--include '*.n??' --exclude '*.fna' --exclude '%s' --exclude '*.log'" % (
safeBaseName)
src = safeDirName + '/'
dest = os.path.join(targetScratchDir, safeBaseName)
blastDb = os.path.join(dest, dest)
rsyncCmd = "rsync -av --omit-dir-times --no-perms %s %s %s" % (rsyncOption,
src, dest)
dbDirNameLocalized, _, exitCode = run_sh_command(rsyncCmd, True, log, True)
if exitCode != 0:
log.error("rsync failed. Cannot localize " + str(db))
return dbDirNameLocalized, RQCExitCodes.JGI_FAILURE
else:
cmd = "chmod -f -R 777 %s" % (dest)
_, _, exitCode = run_sh_command(cmd, True, log)
return blastDb, RQCExitCodes.JGI_SUCCESS
def read_megablast_hits(db, log):
currentDir = RQCReadQcConfig.CFG["output_path"]
megablastDir = "megablast"
megablastPath = os.path.join(currentDir, megablastDir)
statsFile = RQCReadQcConfig.CFG["stats_file"]
filesFile = RQCReadQcConfig.CFG["files_file"]
##
## Process blast output files
##
matchings = 0
hitCount = 0
parsedFile = os.path.join(megablastPath, "megablast.*.%s*.parsed" % (db))
matchings, _, exitCode = run_sh_command(
"grep -v '^#' %s 2>/dev/null | wc -l " % (parsedFile), True, log)
if exitCode == 0: ## if parsed file found.
t = matchings.split()
if len(t) == 1 and t[0].isdigit():
hitCount = int(t[0])
append_rqc_stats(statsFile,
ReadqcStats.ILLUMINA_READ_MATCHING_HITS + " " + db,
hitCount, log)
##
## add .parsed file
##
parsedFileFound, _, exitCode = run_sh_command("ls %s" % (parsedFile),
True, log)
if parsedFileFound:
parsedFileFound = parsedFileFound.strip()
append_rqc_file(filesFile,
ReadqcStats.ILLUMINA_READ_PARSED_FILE + " " + db,
os.path.join(megablastPath, parsedFileFound), log)
else:
log.error("- Failed to add megablast parsed file of %s." % (db))
return RQCExitCodes.JGI_FAILURE
##
## wc the top hits
##
topHit = 0
tophitFile = os.path.join(megablastPath,
"megablast.*.%s*.parsed.tophit" % (db))
tophits, _, exitCode = run_sh_command(
"grep -v '^#' %s 2>/dev/null | wc -l " % (tophitFile), True, log)
t = tophits.split()
if len(t) == 1 and t[0].isdigit():
topHit = int(t[0])
append_rqc_stats(statsFile,
ReadqcStats.ILLUMINA_READ_TOP_HITS + " " + db, topHit,
log)
##
## wc the taxonomic species
##
spe = 0
taxlistFile = os.path.join(megablastPath,
"megablast.*.%s*.parsed.taxlist" % (db))
species, _, exitCode = run_sh_command(
"grep -v '^#' %s 2>/dev/null | wc -l " % (taxlistFile), True, log)
t = species.split()
if len(t) == 1 and t[0].isdigit():
spe = int(t[0])
append_rqc_stats(statsFile,
ReadqcStats.ILLUMINA_READ_TAX_SPECIES + " " + db, spe,
log)
##
## wc the top 100 hit
##
top100hits = 0
top100hitFile = os.path.join(megablastPath,
"megablast.*.%s*.parsed.top100hit" % (db))
species, _, exitCode = run_sh_command(
"grep -v '^#' %s 2>/dev/null | wc -l " % (top100hitFile), True,
log)
t = species.split()
if len(t) == 1 and t[0].isdigit():
top100hits = int(t[0])
append_rqc_stats(statsFile,
ReadqcStats.ILLUMINA_READ_TOP_100HITS + " " + db,
top100hits, log)
##
## Find and add taxlist file
##
taxListFound, _, exitCode = run_sh_command("ls %s" % (taxlistFile),
True, log)
taxListFound = taxListFound.strip()
if taxListFound:
append_rqc_file(filesFile,
ReadqcStats.ILLUMINA_READ_TAXLIST_FILE + " " + db,
os.path.join(megablastPath, taxListFound), log)
else:
log.error("- Failed to add megablast taxlist file of %s." % (db))
return RQCExitCodes.JGI_FAILURE
##
## Find and add tophit file
##
tophitFound, _, exitCode = run_sh_command("ls %s" % (tophitFile), True,
log)
tophitFound = tophitFound.strip()
if tophitFound:
append_rqc_file(filesFile,
ReadqcStats.ILLUMINA_READ_TOPHIT_FILE + " " + db,
os.path.join(megablastPath, tophitFound), log)
else:
log.error("- Failed to add megablast tophit file of %s." % (db))
return RQCExitCodes.JGI_FAILURE
##
## Find and add top100hit file
##
top100hitFound, _, exitCode = run_sh_command("ls %s" % (top100hitFile),
True, log)
top100hitFound = top100hitFound.strip()
if top100hitFound:
append_rqc_file(
filesFile, ReadqcStats.ILLUMINA_READ_TOP100HIT_FILE + " " + db,
os.path.join(megablastPath, top100hitFound), log)
else:
log.error("- Failed to add megablast top100hit file of %s." % (db))
return RQCExitCodes.JGI_FAILURE
else:
log.info("- No blast hits for %s." % (db))
return RQCExitCodes.JGI_SUCCESS
def post_process(fastq, outDir, filteredFastq, status, log):
## obtain read counts from input and filtered fastq files and save the values to STATS_LIST_FILE_NAME file;
## compress the filtered fastq file
log_and_print(
"\n\n%s - RUN POST PROCESS <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<%s\n"
% (color['pink'], color['']))
if STEP_ORDER[status] < STEP_ORDER[POST_END]:
checkpoint(POST_START, status)
rawCnt = 0
rawBaseCnt = 0
newCnt = 0
newBaseCnt = 0
stats = get_dict_obj(BB_STATS_LIST_FILE_NAME)
rawCnt = pipeline_val('inputReads', {
'type': 'int',
'vtype': 'numeric'
}, stats)
rawBaseCnt = pipeline_val('inputBases', {
'type': 'int',
'vtype': 'numeric'
}, stats)
newCnt = pipeline_val('outputReads', {
'type': 'int',
'vtype': 'numeric'
}, stats)
newBaseCnt = pipeline_val('outputBases', {
'type': 'int',
'vtype': 'numeric'
}, stats)
readCounts = {}
readRmPct = 100.0 * ((rawCnt - newCnt) / float(rawCnt))
baseRmPct = 100.0 * ((rawBaseCnt - newBaseCnt) / float(rawBaseCnt))
readCounts['readRmPct'] = '%.3f' % readRmPct
readCounts['baseRmPct'] = '%.3f' % baseRmPct
refStats = {}
filterLogStat = {}
cardinality = None
bbdukVersion = None
bbmapVersion = None
if os.path.isfile("filter.log"):
with open(os.path.join(outDir, "filter.log"), "r") as FLFH:
isContamNumChecked = False ## Contamination will be done twice for removeribo or for MTF
isKtrimmedTotalRemovedNumChecked = False ## for parsing "Total Removed" after ktrimming
for l in FLFH:
if l.startswith("Input:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 2
if 'adaptertriminput' not in filterLogStat:
filterLogStat["adaptertriminput"] = {
"numreads": toks[0],
"numbases": toks[1]
}
elif 'contamtriminput' not in filterLogStat:
filterLogStat["contamtriminput"] = {
"numreads": toks[0],
"numbases": toks[1]
}
elif l.startswith("FTrimmed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["ftrimmed"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("KTrimmed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["ktrimmed"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
isKtrimmedTotalRemovedNumChecked = True
## RQCSUPPORT-1987
elif l.startswith("Total Removed:"
) and isKtrimmedTotalRemovedNumChecked:
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["ktrimmed_total_removed"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
isKtrimmedTotalRemovedNumChecked = False
elif l.startswith("Trimmed by overlap:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["trimmedbyoverlap"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("Result:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
if 'adaptertrimresult' not in filterLogStat:
filterLogStat["adaptertrimresult"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif 'contamtrimresult' not in filterLogStat:
filterLogStat["contamtrimresult"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("Unique 31-mers:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 2 or len(toks) == 1
if 'adaptertrimunique31mers' not in filterLogStat:
if len(toks) == 2:
filterLogStat["adaptertrimunique31mers"] = {
"num": toks[1]
}
else:
filterLogStat["adaptertrimunique31mers"] = {
"num": "0"
}
else:
if len(toks) == 2:
filterLogStat["contamtrimunique31mers"] = {
"num": toks[1]
}
else:
filterLogStat["contamtrimunique31mers"] = {
"num": "0"
}
elif not isContamNumChecked and l.startswith(
"Contaminants:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["contaminants"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
isContamNumChecked = True
elif l.startswith("QTrimmed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["qtrimmed"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("Short Read Discards:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["shortreaddiscards"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("Low quality discards:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["lowqualitydiscards"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("BBDuk version"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 1
bbdukVersion = toks[0]
elif l.startswith("BBMap version"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 1
bbmapVersion = toks[0]
## BBDuk 36.12 06272016
elif l.startswith("Adapter Sequence Removed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["adaptersequenceremoved"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("Synthetic Contam Sequence Removed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["syntheticcontamsequenceremoved"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
## 08112016
elif l.startswith(
"Short Synthetic Contam Sequence Removed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat[
"shortsyntheticcontamsequenceremoved"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
elif l.startswith("Ribosomal Sequence Removed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["ribosomalsequenceremoved"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
## BBMap 36.12 06272016
elif l.startswith("Human Sequence Removed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["humansequenceremoved"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
## RQC-862, RQC-880
elif l.startswith("Microbial Sequence Removed:"):
toks = re.findall("(\d+.\d*)", l.rstrip())
assert len(toks) == 4
filterLogStat["microbialremoved"] = {
"numreads": toks[0],
"percreads": toks[1],
"numbases": toks[2],
"percbases": toks[3]
}
##
## refStats.txt format
##
## name %unambiguousReads unambiguousMB %ambiguousReads ambiguousMB unambiguousReads ambiguousReads
## human_masked 85.24693 498.92052 0.09378 0.55290 3350692 3686
## mouse_masked 0.03765 0.21670 0.10802 0.63690 1480 4246
## cat_masked 0.01862 0.09568 0.02514 0.14820 732 988
## dog_masked 0.00697 0.03815 0.01384 0.08160 274 544
##
if os.path.isfile("refStats.txt"):
refStatsFile = os.path.join(outDir, "refStats.txt")
with open(refStatsFile) as RFH:
## Need to report 0 if nothing matched
refStats['human'] = {
"unambiguousReadsPerc": "0",
"unambiguousMB": "0",
"ambiguousReadsPerc": "0",
"ambiguousMB": "0",
"unambiguousReads": "0",
"ambiguousReads": "0",
"totalPerc": "0"
}
refStats['cat'] = {
"unambiguousReadsPerc": "0",
"unambiguousMB": "0",
"ambiguousReadsPerc": "0",
"ambiguousMB": "0",
"unambiguousReads": "0",
"ambiguousReads": "0",
"totalPerc": "0"
}
refStats['dog'] = {
"unambiguousReadsPerc": "0",
"unambiguousMB": "0",
"ambiguousReadsPerc": "0",
"ambiguousMB": "0",
"unambiguousReads": "0",
"ambiguousReads": "0",
"totalPerc": "0"
}
refStats['mouse'] = {
"unambiguousReadsPerc": "0",
"unambiguousMB": "0",
"ambiguousReadsPerc": "0",
"ambiguousMB": "0",
"unambiguousReads": "0",
"ambiguousReads": "0",
"totalPerc": "0"
}
for l in RFH:
if l:
if l.startswith("#"):
continue
toks = l.rstrip().split()
assert len(toks) == 7
## the number and percent of reads that map unambiguously or ambiguously to human, cat, dog.
## take the sum of the two numbers (ambiguous plus unambiguous) to use as the final percentage.
if l.startswith("human"):
refStats['human'] = {
"unambiguousReadsPerc": toks[1],
"unambiguousMB": toks[2],
"ambiguousReadsPerc": toks[3],
"ambiguousMB": toks[4],
"unambiguousReads": toks[5],
"ambiguousReads": toks[6],
"totalPerc": float(toks[3]) + float(toks[1])
}
if l.startswith("cat"):
refStats['cat'] = {
"unambiguousReadsPerc": toks[1],
"unambiguousMB": toks[2],
"ambiguousReadsPerc": toks[3],
"ambiguousMB": toks[4],
"unambiguousReads": toks[5],
"ambiguousReads": toks[6],
"totalPerc": float(toks[3]) + float(toks[1])
}
if l.startswith("dog"):
refStats['dog'] = {
"unambiguousReadsPerc": toks[1],
"unambiguousMB": toks[2],
"ambiguousReadsPerc": toks[3],
"ambiguousMB": toks[4],
"unambiguousReads": toks[5],
"ambiguousReads": toks[6],
"totalPerc": float(toks[3]) + float(toks[1])
}
if l.startswith("mouse"):
refStats['mouse'] = {
"unambiguousReadsPerc": toks[1],
"unambiguousMB": toks[2],
"ambiguousReadsPerc": toks[3],
"ambiguousMB": toks[4],
"unambiguousReads": toks[5],
"ambiguousReads": toks[6],
"totalPerc": float(toks[3]) + float(toks[1])
}
log.debug("refStats.txt: %s", str(refStats))
###########################################################
log_and_print("Write to stats file %s" % STATS_LIST_FILE_NAME)
###########################################################
if os.path.isfile(STATS_LIST_FILE_NAME):
os.remove(STATS_LIST_FILE_NAME)
with open(BB_STATS_LIST_FILE_NAME) as bbfh:
with open(STATS_LIST_FILE_NAME, 'a') as fh:
for line in bbfh:
if not line.startswith("#") and line.strip():
fh.write(line)
bbtoolsVersion = None
stats = get_dict_obj(STATS_LIST_FILE_NAME)
with open(STATS_LIST_FILE_NAME, 'a') as fh:
for key in readCounts:
if key not in stats:
write_stats(fh, key, readCounts[key], log)
for key in refStats:
for k in refStats[key]:
write_stats(fh, key + '_' + k, refStats[key][k], log)
write_stats(fh, "cardinality", cardinality, log)
## Write refStats to filterStats.txt file
for key in filterLogStat:
for k in filterLogStat[key]:
write_stats(fh, key + '_' + k, filterLogStat[key][k], log)
bbversionCmd = os.path.join(BBDIR, 'bbversion.sh')
cmd = "%s" % (bbversionCmd)
stdOut, _, exitCode = run_sh_command(cmd, True, log)
assert stdOut is not None
bbtoolsVersion = stdOut.strip()
## 05112017 Now bbtools version = bbmap version
# bbtoolsVersion = bbmapVersion if bbmapVersion else "37.xx"
assert bbtoolsVersion is not None
write_stats(fh, "filter_tool", "bbtools " + bbtoolsVersion, log)
write_stats(fh, "filter", VERSION, log)
## Version recording
if bbdukVersion is None: bbdukVersion = bbtoolsVersion
if bbmapVersion is None: bbmapVersion = bbtoolsVersion
write_stats(fh, "bbduk_version", bbdukVersion, log)
write_stats(fh, "bbmap_version", bbmapVersion, log)
checkpoint(POST_END, status)
status = POST_END
else:
log_and_print('No need to do post processing.')
return filteredFastq, status
def run_rqcfilter(infastq, outDir, prodType, status, enableRmoveMicrobes,
enableAggressive, disableRmoveMicrobes, disableClumpify,
taxList, rdb, log):
log_and_print(
"\n\n%s - RUN RQCFILTER <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<%s\n"
% (color['pink'], color['']))
make_dir_p(outDir)
opt = None
extraOptions = ""
optionFile = ""
if prodType.endswith("-OLD"):
extraOptions = " barcodefilter=f chastityfilter=f"
prodType = prodType.replace("-OLD", "")
if prodType in FILTER_METHODS_TXT:
optionFile = os.path.join(
SRC_ROOT, "filter_param/",
FILTER_METHODS_TXT[prodType].replace(".txt", ".config"))
log_and_print("Filter option file: %s" % optionFile)
opt = open(optionFile, 'r').readline().rstrip()
else:
log_and_print("The product type, %s, is not supported yet." % prodType)
sys.exit(2)
assert (opt), "Null filter options."
if prodType in ("METATRANSCRIPTOME", "MTF"):
if infastq.endswith(".gz"):
opt += " outribo=%s " % (os.path.basename(infastq).replace(
".fastq", ".rRNA.fastq"))
else:
opt += " outribo=%s " % (os.path.basename(infastq).replace(
".fastq", ".rRNA.fastq.gz"))
if enableRmoveMicrobes:
if opt.find("removemicrobes=f") != -1:
opt = opt.replace("removemicrobes=f", "removemicrobes=t")
opt += " removemicrobes=t "
if disableRmoveMicrobes:
if opt.find("removemicrobes=t") != -1:
opt = opt.replace("removemicrobes=t", "removemicrobes=f")
else:
opt += " removemicrobes=f "
if enableAggressive:
opt += " aggressive=t microbebuild=3 "
if taxList:
opt += " taxlist=%s " % (taxList)
## Temp set clumpify=t for all prod types (RQC-890)
if not disableClumpify:
opt += " clumpify=t "
else:
opt += " clumpify=f "
opt += " tmpdir=null "
opt += extraOptions
cmd = os.path.join(BBDIR, "rqcfilter.sh")
filterLogFile = os.path.join(outDir, "filter.log")
cmdStr = "%s in=%s path=%s %s usejni=f rqcfilterdata=%s > %s 2>&1" % (
cmd, infastq, outDir, opt, rdb, filterLogFile)
rtn = [None, status]
outFastqFile = None
shFileName = "%s/filter.sh" % outDir
def find_filtered_fastq(adir):
outFastqFile = None
searchPatt = os.path.join(adir, "*.fastq.gz")
outFastqFileList = glob.glob(searchPatt)
assert len(outFastqFileList
) >= 1, "ERROR: cannot find *.fastq.gz output file."
for f in outFastqFileList:
f = os.path.basename(f)
t = f.split(".")
if t[-3] not in ("frag", "singleton", "unknown", "rRNA", "lmp"):
filterCode = t[-3]
elif t[-3] == "lmp": ## nextera
filterCode = t[-4]
if len(t) == 7: ## ex) 12345.1.1234.ACCCC.anqdpht.fastq.gz
fileNamePrefix = '.'.join(t[:4])
elif len(t) == 6: ## ex) 6176.5.39297.anqrpht.fastq.gz
fileNamePrefix = '.'.join(t[:3])
else:
log.warning("Unexpected filtered file name, %s",
outFastqFileList)
fileNamePrefix = '.'.join(t[:-3])
log_and_print("Use %s as file prefix." % fileNamePrefix)
assert filterCode and fileNamePrefix, "ERROR: unexpected filter file name: %s" % (
outFastqFileList)
of = os.path.join(adir,
'.'.join([fileNamePrefix, filterCode, "fastq.gz"]))
lof = os.path.join(
adir, '.'.join([fileNamePrefix, filterCode, "lmp.fastq.gz"]))
if os.path.isfile(of):
outFastqFile = of
elif os.path.isfile(lof):
outFastqFile = lof
else:
log.error("Cannot find fastq.gz file.")
# rename output file to *.filtered.fastq.gz
f = os.path.basename(outFastqFile)
t = f.split(".")
if t[-3] != 'filtered':
fto = os.path.join(
adir, '.'.join(['.'.join(t[:-3]), 'filtered', "fastq.gz"]))
shutil.move(outFastqFile, fto)
outFastqFile = fto
return outFastqFile
def find_filter_number(outFastqFile):
filteredReadNum = fastqUtil.check_fastq_format(outFastqFile) / 4
if filteredReadNum < 0:
log_and_print("RUN RQCFILTER - filtered fastq format error: %s." %
outFastqFile)
return filteredReadNum
if STEP_ORDER[status] < STEP_ORDER[RQCFILTER_END]:
create_shell(shFileName, (cmdStr, ))
log_and_print("rqcfilter cmd=[%s]" % cmdStr)
log_and_print("sh file name=[%s]" % shFileName)
stdOut, stdErr, exitCode = run_sh_command(
shFileName, True, log, True) ## stdOut of 0 is success
if exitCode != 0:
log.error("Failed to run : %s, stdout : %s, stderr: %s",
shFileName, stdOut, stdErr)
return rtn
outFastqFile = find_filtered_fastq(outDir)
filteredReadNum = find_filter_number(outFastqFile)
log_and_print("Read counts after RQCFILTER step = %d" %
filteredReadNum)
log_and_print("RUN RQCFILTER - completed")
checkpoint(RQCFILTER_END, status)
status = RQCFILTER_END
if filteredReadNum == 0:
log.warning("No reads left after filtering")
checkpoint(PIPE_COMPLETE, status)
with open(BB_STATS_LIST_FILE_NAME, 'a') as fh:
write_stats(fh, FILTER_READ_COUNT, 0, log)
write_stats(fh, FILTER_READ_BASE_COUNT, 0, log)
else:
log_and_print(
"No need to rerun RQCFILTER step, get filtered files and stats ... "
)
outFastqFile = find_filtered_fastq(outDir)
rtn = [outFastqFile, status]
return rtn
