def fromsra(args):
"""
%prog fromsra srafile
Convert sra file to fastq using the sratoolkit `fastq-dump`
"""
p = OptionParser(fromsra.__doc__)
p.add_option(
"--paired",
default=False,
action="store_true",
help="Specify if library layout is paired-end",
)
p.add_option(
"--compress",
default=None,
choices=["gzip", "bzip2"],
help="Compress output fastq files",
)
p.set_outdir()
p.set_grid()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
(srafile, ) = args
paired = opts.paired
compress = opts.compress
outdir = opts.outdir
script_path = which("fastq-dump")
if not script_path:
logging.error("Cannot find `fastq-dump` in the PATH")
sys.exit()
cmd = [script_path]
if compress:
cmd.append("--{0}".format(compress))
if paired:
cmd.append("--split-files")
if outdir:
cmd.append("--outdir {0}".format(outdir))
cmd.append(srafile)
outcmd = " ".join(cmd)
sh(outcmd, grid=opts.grid)
def fromsra(args):
"""
%prog fromsra srafile
Convert sra file to fastq using the sratoolkit `fastq-dump`
"""
p = OptionParser(fromsra.__doc__)
p.add_option(
"--paired",
default=False,
action="store_true",
help="Specify if library layout is paired-end " + "[default: %default]",
)
p.add_option(
"--compress", default=None, choices=["gzip", "bzip2"], help="Compress output fastq files [default: %default]"
)
p.set_outdir()
p.set_grid()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
srafile, = args
paired = opts.paired
compress = opts.compress
outdir = opts.outdir
script_path = which("fastq-dump")
if not script_path:
logging.error("Cannot find `fastq-dump` in the PATH")
sys.exit()
cmd = [script_path]
if compress:
cmd.append("--{0}".format(compress))
if paired:
cmd.append("--split-3")
if outdir:
cmd.append("--outdir {0}".format(outdir))
cmd.append(srafile)
outcmd = " ".join(cmd)
sh(outcmd, grid=opts.grid)
def assemble(args):
"""
%prog assemble pasa_db_name genome.fasta transcripts-dn.fasta [transcript-gg.fasta]
Run the PASA alignment assembly pipeline
If two transcript fasta files (Trinity denovo and genome guided) are provided
and the `--compreh` param is enabled, the PASA Comprehensive Transcriptome DB
protocol is followed <http://pasa.sourceforge.net/#A_ComprehensiveTranscriptome>
Using the `--prepare` option creates a shell script with the run commands without
executing the pipeline
"""
p = OptionParser(assemble.__doc__)
p.set_pasa_opts()
p.add_option("--prepare", default=False, action="store_true",
help="Prepare PASA run script with commands [default: %default]")
p.set_grid()
p.set_grid_opts()
opts, args = p.parse_args(args)
if len(args) not in (3, 4):
sys.exit(not p.print_help())
pasa_db, genome, dnfasta, = args[:3]
ggfasta = args[3] if len(args) == 4 else None
PASA_HOME = opts.pasa_home
if not op.isdir(PASA_HOME):
logging.error("PASA_HOME={0} directory does not exist".format(PASA_HOME))
sys.exit()
aligners = opts.aligners.split(",")
for aligner in aligners:
if aligner not in ALLOWED_ALIGNERS:
logging.error("Error: Unknown aligner `{0}`".format(aligner))
logging.error("Can be any of {0}, ".format("|".join(ALLOWED_ALIGNERS)) + \
"combine multiple aligners in list separated by comma")
sys.exit()
clean = opts.clean
seqclean = op.join(opts.tgi_home, "seqclean")
accn_extract = which(op.join(PASA_HOME, "misc_utilities", \
"accession_extractor.pl"))
launch_pasa = which(op.join(PASA_HOME, "scripts", \
"Launch_PASA_pipeline.pl"))
build_compreh_trans = which(op.join(PASA_HOME, "scripts", \
"build_comprehensive_transcriptome.dbi"))
fl_accs = opts.fl_accs
cpus = opts.cpus
grid = opts.grid
prepare, runfile = opts.prepare, "run.sh"
pctcov, pctid = opts.pctcov, opts.pctid
compreh_pctid = opts.compreh_pctid
compreh_pctcov, bpsplice = opts.compreh_pctcov, opts.bpsplice
cmds = []
# set PASAHOME env variable if preparing shell script
if prepare:
env_cmd = 'export PASAHOME="{0}"'.format(PASA_HOME)
cmds.append(env_cmd)
if ggfasta:
transcripts = FileMerger([dnfasta, ggfasta], tfasta).merge()
accn_extract_cmd = "cat {0} | {1} > {2}".format(dnfasta, accn_extract, tdn)
cmds.append(accn_extract_cmd)
if not prepare:
sh(accn_extract_cmd)
else:
symlink(dnfasta, tfasta)
transcripts = tfasta
if opts.grid and not opts.threaded:
opts.threaded = opts.cpus
prjobid = None
if clean:
ccpus = 16 if cpus >= 16 else cpus
cleancmd = "{0} {1} -c {2} -l 60".format(seqclean, transcripts, ccpus)
if prepare:
cmds.append(cleancmd)
else:
prjobid = sh(cleancmd, grid=grid, grid_opts=opts)
aafw = must_open(aaconf, "w")
print(alignAssembly_conf.format("{0}_pasa".format(pasa_db), \
pctcov, pctid, bpsplice), file=aafw)
aafw.close()
symlink(genome, gfasta)
aacmd = "{0} -c {1} -C -R -g {2}".format(launch_pasa, aaconf, gfasta)
aacmd += " -t {0}.clean -T -u {0}".format(transcripts) if clean else \
" -t {0}".format(transcripts)
if fl_accs:
symlink(fl_accs, flaccs)
aacmd += " -f {0}".format(flaccs)
if ggfasta:
aacmd += " --TDN {0}".format(tdn)
aacmd += " --ALIGNERS {0} -I {1} --CPU {2}".format(",".join(aligners), \
opts.intron, cpus)
if prepare:
cmds.append(aacmd)
else:
opts.hold_jid = prjobid
prjobid = sh(aacmd, grid=grid, grid_opts=opts)
if opts.compreh and ggfasta:
comprehcmd = "{0} -c {1} -t {2}".format(build_compreh_trans, aaconf, transcripts)
comprehcmd += " --min_per_ID {0} --min_per_aligned {1}".format(compreh_pctid, compreh_pctcov)
if prepare:
cmds.append(comprehcmd)
else:
opts.hold_jid = prjobid
prjobid = sh(comprehcmd, grid=grid, grid_opts=opts)
if prepare:
write_file(runfile, "\n".join(cmds)) # initialize run script
def compare(args):
"""
%prog compare pasa_db_name [--annots_gff3=annotation.gff3]
Run the PASA annotation comparison pipeline
This assumes that PASA alignment assembly has alredy been completed and
run directory contains `genome.fasta` and `transcript.fasta` files.
If `--annots_gff3` is specified, the PASA database is loaded with the annotations
first before starting annotation comparison. Otherwise, it uses previously
loaded annotation data.
Using the `--prepare` option creates a shell script with the run commands without
executing the pipeline
"""
p = OptionParser(compare.__doc__)
p.set_pasa_opts(action="compare")
p.add_option("--prepare", default=False, action="store_true",
help="Prepare PASA run script with commands [default: %default]")
p.set_grid()
p.set_grid_opts()
opts, args = p.parse_args(args)
if len(args) < 1:
sys.exit(not p.print_help())
pasa_db, = args
PASA_HOME = opts.pasa_home
if not op.isdir(PASA_HOME):
logging.error("PASA_HOME={0} directory does not exist".format(PASA_HOME))
sys.exit()
launch_pasa = which(op.join(PASA_HOME, "scripts", \
"Launch_PASA_pipeline.pl"))
annots_gff3 = opts.annots_gff3
grid = opts.grid
prepare, runfile = opts.prepare, "run.sh"
os.chdir(pasa_db)
if prepare:
write_file(runfile, "", append=True, skipcheck=True) # initialize run script
acfw = must_open(acconf, "w")
print(annotCompare_conf.format("{0}_pasa".format(pasa_db), \
opts.pctovl, opts.pct_coding, opts.pctid_prot, opts.pctlen_FL, \
opts.pctlen_nonFL, opts.orf_size, opts.pct_aln, opts.pctovl_gene, \
opts.stompovl, opts.trust_FL, opts.utr_exons), file=acfw)
acfw.close()
if not op.exists(gfasta):
sys.exit("Genome fasta file `{0}` does not exist".format(gfasta))
transcripts = tfasta
if not op.exists(transcripts):
sys.exit("Transcript fasta file `{0}` does not exist".format(transcripts))
if op.exists("{0}.clean".format(transcripts)):
transcripts = "{0}.clean".format(transcripts)
accmd = "{0} -c {1} -A -g {2} -t {3} --GENETIC_CODE {4}".format(launch_pasa, \
acconf, gfasta, transcripts, opts.genetic_code)
if annots_gff3:
if not op.exists(annots_gff3):
sys.exit("Annotation gff3 file `{0}` does not exist".format(annots_gff3))
symlink(annots_gff3, annotation)
accmd += " -L --annots_gff3 {0}".format(annotation)
if prepare:
write_file(runfile, accmd, append=True)
else:
sh(accmd, grid=grid, grid_opts=opts)
def prepare(args):
"""
%prog prepare [--options] folder [genome.fasta]
Run Trinity on a folder of reads. When paired-end (--paired) mode is on,
filenames will be scanned based on whether they contain the patterns
("_1_" and "_2_") or (".1." and ".2.") or ("_1." and "_2.").
By default, prepare script for DN
If genome.fasta is provided, prepare script for GG-Trinity.
If coord-sorted BAM is provided, then it will use it as starting point.
Since GG-Trinity jobs are partitioned DN-Trinity jobs run on relatively small
regions, lesser amount of CPU can be specified for each DN job using `--gg_cpu`
In such cases, the `--cpu` should be set to a larger value to help speedup
upstream steps such as GSNAP read mapping or coordinate sorting of BAM files.
Newer versions of trinity can take multiple fastq files as input.
If "--merge" is specified, the fastq files are merged together before assembling
"""
p = OptionParser(prepare.__doc__)
p.add_option("--paired", default=False, action="store_true",
help="Paired-end mode [default: %default]")
p.add_option("--merge", default=False, action="store_true",
help="Merge individual input fastq's into left/right/single" + \
" file(s) [default: %default]")
p.set_trinity_opts()
p.set_grid()
opts, args = p.parse_args(args)
if len(args) not in (1, 2):
sys.exit(not p.print_help())
inparam, = args[:1]
genome = args[1] if len(args) == 2 else None
method = "GG" if genome is not None else "DN"
paired = opts.paired
merge = opts.merge
thome = opts.trinity_home
use_bam = opts.use_bam
gg_cpu = opts.gg_cpu
pf = inparam.split(".")[0]
tfolder = "{0}_{1}".format(pf, method)
cwd = os.getcwd()
mkdir(tfolder)
os.chdir(tfolder)
flist = iglob("../" + inparam, "*.fq", "*.fastq", "*.fq.gz", "*.fastq.gz")
if paired:
f1 = [x for x in flist if "_1_" in x or ".1." in x or "_1." in x]
f2 = [x for x in flist if "_2_" in x or ".2." in x or "_2." in x]
assert len(f1) == len(f2)
if merge:
r1, r2 = "left.fastq", "right.fastq"
reads = ((f1, r1), (f2, r2))
else:
if merge:
r = "single.fastq"
reads = ((flist, r), )
if merge:
for fl, r in reads:
fm = FileMerger(fl, r)
fm.merge(checkexists=True)
cmd = op.join(thome, "Trinity")
cmd += " --seqType fq --JM {0} --CPU {1}".format(opts.JM, opts.cpus)
cmd += " --min_contig_length {0}".format(opts.min_contig_length)
if opts.bflyGCThreads:
cmd += " --bflyGCThreads {0}".format(opts.bflyGCThreads)
if method == "GG":
cmd += " --genome {0} --genome_guided_max_intron {1}".format(genome, opts.max_intron)
if use_bam:
cmd += " --genome_guided_use_bam {0}".format(use_bam)
if gg_cpu:
cmd += " --genome_guided_CPU {0}".format(gg_cpu)
if opts.grid and opts.grid_conf_file:
cmd += " --grid_conf_file={0}".format(opts.grid_conf_file)
if paired:
if merge:
cmd += " --left {0} --right {1}".format(reads[0][-1], reads[1][-1])
else:
for lf, rf in zip(f1, f2):
cmd += " --left {0}".format(lf)
cmd += " --right {0}".format(rf)
else:
if merge:
cmd += " --single {0}".format(reads[0][-1])
else:
for f in flist:
cmd += " --single {0}".format(f)
if opts.extra:
cmd += " {0}".format(opts.extra)
runfile = "run.sh"
write_file(runfile, cmd)
os.chdir(cwd)
def prepare(args):
"""
%prog prepare [--options] folder [--bam rnaseq.coordSorted.bam]
Run Trinity on a folder of reads. When paired-end (--paired) mode is on,
filenames will be scanned based on whether they contain the patterns
("_1_" and "_2_") or (".1." and ".2.") or ("_1." and "_2.").
By default, prepare script for DN-Trinity.
If coord-sorted BAM is provided, prepare script for GG-Trinity, using BAM
as starting point.
Newer versions of trinity can take multiple fastq files as input.
If "--merge" is specified, the fastq files are merged together before assembling
"""
p = OptionParser(prepare.__doc__)
p.add_option("--paired", default=False, action="store_true",
help="Paired-end mode [default: %default]")
p.add_option("--merge", default=False, action="store_true",
help="Merge individual input fastq's into left/right/single" + \
" file(s) [default: %default]")
p.set_trinity_opts()
p.set_fastq_names()
p.set_grid()
opts, args = p.parse_args(args)
if len(args) not in (1, 2):
sys.exit(not p.print_help())
inparam, = args[:1]
paired = opts.paired
merge = opts.merge
trinity_home = opts.trinity_home
hpc_grid_runner_home = opts.hpcgridrunner_home
method = "DN"
bam = opts.bam
if bam and op.exists(bam):
bam = op.abspath(bam)
method = "GG"
pf = inparam.split(".")[0]
tfolder = "{0}_{1}".format(pf, method)
cwd = os.getcwd()
mkdir(tfolder)
os.chdir(tfolder)
cmds = []
# set TRINITY_HOME env variable when preparing shell script
env_cmd = 'export TRINITY_HOME="{0}"'.format(trinity_home)
cmds.append(env_cmd)
if method == "DN":
assert op.exists("../" + inparam)
flist = iglob("../" + inparam, opts.names)
if paired:
f1 = [x for x in flist if "_1_" in x or ".1." in x or "_1." in x or "_R1" in x]
f2 = [x for x in flist if "_2_" in x or ".2." in x or "_2." in x or "_R2" in x]
assert len(f1) == len(f2)
if merge:
r1, r2 = "left.fastq", "right.fastq"
reads = ((f1, r1), (f2, r2))
else:
if merge:
r = "single.fastq"
reads = ((flist, r), )
if merge:
for fl, r in reads:
fm = FileMerger(fl, r)
fm.merge(checkexists=True)
cmd = op.join(trinity_home, "Trinity")
cmd += " --seqType fq --max_memory {0} --CPU {1}".format(opts.max_memory, opts.cpus)
cmd += " --min_contig_length {0}".format(opts.min_contig_length)
if opts.bflyGCThreads:
cmd += " --bflyGCThreads {0}".format(opts.bflyGCThreads)
if method == "GG":
cmd += " --genome_guided_bam {0}".format(bam)
cmd += " --genome_guided_max_intron {0}".format(opts.max_intron)
else:
if paired:
if merge:
cmd += " --left {0} --right {1}".format(reads[0][-1], reads[1][-1])
else:
cmd += " --left {0}".format(",".join(f1))
cmd += " --right {0}".format(",".join(f2))
else:
if merge:
cmd += " --single {0}".format(reads[0][-1])
else:
for f in flist:
cmd += " --single {0}".format(f)
if opts.grid and opts.grid_conf_file:
hpc_grid_runner = op.join(hpc_grid_runner_home, "hpc_cmds_GridRunner.pl")
hpc_grid_conf_file = op.join(hpc_grid_runner_home, "hpc_conf", opts.grid_conf_file)
assert op.exists(hpc_grid_conf_file), "HpcGridRunner conf file does not exist: {0}".format(hpc_grid_conf_file)
cmd += ' --grid_exec "{0} --grid_conf {1} -c"'.format(hpc_grid_runner, hpc_grid_conf_file)
if opts.extra:
cmd += " {0}".format(opts.extra)
cmds.append(cmd)
if opts.cleanup:
cleanup_cmd = 'rm -rf !("Trinity.fasta"|"Trinity.gene_trans_map"|"Trinity.timing")' \
if method == "DN" else \
'rm -rf !("Trinity-GG.fasta"|"Trinity-GG.gene_trans_map"|"Trinity.timing")'
cmd.append(cleanup_cmd)
runfile = "run.sh"
write_file(runfile, "\n".join(cmds))
os.chdir(cwd)
def assemble(args):
"""
%prog assemble pasa_db_name genome.fasta transcripts-dn.fasta [transcript-gg.fasta]
Run the PASA alignment assembly pipeline
If two transcript fasta files (Trinity denovo and genome guided) are provided,
the PASA Comprehensive Transcriptome protocol is followed
<http://pasa.sourceforge.net/#A_ComprehensiveTranscriptome>
Using the `--prepare` option creates a shell script with the run commands without
executing the pipeline
"""
p = OptionParser(assemble.__doc__)
p.set_home("pasa")
p.set_align(pctid=95, pctcov=90, intron=2000, bpsplice=3, compreh_pctcov=30)
p.add_option("--aligners", default="blat,gmap",
help="Specify splice aligners to use for mapping [default: %default]")
p.add_option("--clean", default=False, action="store_true",
help="Clean transcripts using tgi seqclean [default: %default]")
p.set_cpus()
p.set_grid()
p.set_grid_opts()
p.add_option("--prepare", default=False, action="store_true",
help="Prepare PASA run script with commands [default: %default]")
opts, args = p.parse_args(args)
if len(args) not in (3, 4):
sys.exit(not p.print_help())
pasa_db, genome, dnfasta, = args[:3]
ggfasta = args[3] if len(args) == 4 else None
PASA_HOME = opts.pasa_home
if not op.isdir(PASA_HOME):
logging.error("PASA_HOME={0} directory does not exist".format(PASA_HOME))
sys.exit()
aligners = opts.aligners.split(",")
for aligner in aligners:
if aligner not in ALLOWED_ALIGNERS:
logging.error("Error: Unknown aligner `{0}`".format(aligner))
logging.error("Can be any of {0}, ".format("|".join(ALLOWED_ALIGNERS)) + \
"combine multiple aligners in list separated by comma")
sys.exit()
clean = opts.clean
seqclean = which("seqclean")
if clean and not seqclean:
logging.error("Cannot find tgi seqclean in PATH")
sys.exit()
accn_extract = which(op.join(PASA_HOME, "misc_utilities", "accession_extractor.pl"))
launch_pasa = which(op.join(PASA_HOME, "scripts", "Launch_PASA_pipeline.pl"))
build_compreh_trans = which(op.join(PASA_HOME, "scripts", "build_comprehensive_transcriptome.dbi"))
cpus = opts.cpus
grid = opts.grid
prepare, runfile = opts.prepare, "run.sh"
pctcov, pctid = opts.pctcov, opts.pctid
compreh_pctcov, bpsplice = opts.compreh_pctcov, opts.bpsplice
mkdir(pasa_db)
os.chdir(pasa_db)
if prepare:
write_file(runfile, "") # initialize run script
if ggfasta:
transcripts = FileMerger([dnfasta, ggfasta], tfasta).merge()
accn_extract_cmd = "cat {0} | {1} > {2}".format(dnfasta, accn_extract, tdn)
write_file(runfile, accn_extract_cmd, append=True) \
if prepare else sh(accn_extract_cmd)
else:
transcripts = dnfasta
if opts.grid and not opts.threaded:
opts.threaded = opts.cpus
prjobid = None
if clean:
cleancmd = "{0} {1} -c {2} -l 60".format(seqclean, transcripts, cpus)
if prepare:
write_file(runfile, cleancmd, append=True)
else:
prjobid = sh(cleancmd, grid=grid, grid_opts=opts)
aafw = must_open(aaconf, "w")
print >> aafw, alignAssembly_conf.format("{0}_pasa".format(pasa_db), pctcov, pctid, bpsplice)
aafw.close()
aacmd = "{0} -c {1} -C -R -g {2}".format(launch_pasa, aaconf, genome)
aacmd += " -t {0}.clean -T -u {0} ".format(transcripts) if clean else \
" -t {0} ".format(transcripts)
if ggfasta:
aacmd += " --TDN {0} ".format(tdn)
aacmd += " --ALIGNERS {0} -I {1}".format(",".join(aligners), opts.intron)
if prepare:
write_file(runfile, aacmd, append=True)
else:
opts.hold_jid = prjobid
prjobid = sh(aacmd, grid=grid, grid_opts=opts)
if ggfasta:
comprehcmd = "{0} -c {1} -t {2}".format(build_compreh_trans, aaconf, transcripts)
comprehcmd += "--min_per_ID {0} --min_per_aligned {1}".format(pctid, pctcov)
if prepare:
write_file(runfile, comprehcmd, append=True)
else:
opts.hold_jid = prjobid
prjobid = sh(comprehcmd, grid=grid, grid_opts=opts)
def compare(args):
"""
%prog compare pasa_db_name genome.fasta transcripts.fasta [annotation.gff]
Run the PASA annotation comparison pipeline
If annotation.gff file is provided, the PASA database is loaded with the annotations
first before starting annotation comparison. Otherwise, it uses previously
loaded annotation data.
Using the `--prepare` option creates a shell script with the run commands without
executing the pipeline
"""
p = OptionParser(compare.__doc__)
p.set_pasa_opts(action="compare")
p.add_option("--prepare", default=False, action="store_true",
help="Prepare PASA run script with commands [default: %default]")
p.set_grid()
p.set_grid_opts()
opts, args = p.parse_args(args)
if len(args) not in (3, 4):
sys.exit(not p.print_help())
pasa_db, genome, transcripts, = args[:3]
annotation = args[3] if len(args) == 4 else None
PASA_HOME = opts.pasa_home
if not op.isdir(PASA_HOME):
logging.error("PASA_HOME={0} directory does not exist".format(PASA_HOME))
sys.exit()
launch_pasa = which(op.join(PASA_HOME, "scripts", \
"Launch_PASA_pipeline.pl"))
grid = opts.grid
prepare, runfile = opts.prepare, "run.sh"
os.chdir(pasa_db)
if prepare:
write_file(runfile, "") # initialize run script
if opts.grid and not opts.threaded:
opts.threaded = opts.cpus
acfw = must_open(acconf, "w")
print >> acfw, annotCompare_conf.format("{0}_pasa".format(pasa_db), \
opts.pctovl, opts.pct_coding, opts.pctid_prot, opts.pctlen_FL, \
opts.pctlen_nonFL, opts.orf_size, opts.pct_aln, opts.pctovl_gene, \
opts.stompovl, opts.trust_FL, opts.utr_exons)
acfw.close()
if op.exists("{0}.clean".format(transcripts)):
transcripts = "{0}.clean".format(transcripts)
accmd = "{0} -c {1} -A -g {2} -t {3} --GENETIC_CODE {4}".format(launch_pasa, \
acconf, genome, transcripts, opts.genetic_code)
if annotation:
accmd += " -L --annots_gff3 {0}".format(annotation)
if prepare:
write_file(runfile, accmd, append=True)
else:
sh(accmd, grid=grid, grid_opts=opts)
def prepare(args):
"""
%prog prepare [--options] folder [genome.fasta]
Run Trinity on a folder of reads. When paired-end (--paired) mode is on,
filenames will be scanned based on whether they contain the patterns
("_1_" and "_2_") or (".1." and ".2.") or ("_1." and "_2.").
By default, prepare script for DN
If genome.fasta is provided, prepare script for GG-Trinity.
If coord-sorted BAM is provided, then it will use it as starting point.
Since GG-Trinity jobs are partitioned DN-Trinity jobs run on relatively small
regions, lesser amount of CPU can be specified for each DN job using `--gg_cpu`
In such cases, the `--cpu` should be set to a larger value to help speedup
upstream steps such as GSNAP read mapping or coordinate sorting of BAM files.
Newer versions of trinity can take multiple fastq files as input.
If "--merge" is specified, the fastq files are merged together before assembling
"""
p = OptionParser(prepare.__doc__)
p.add_option("--paired",
default=False,
action="store_true",
help="Paired-end mode [default: %default]")
p.add_option("--merge", default=False, action="store_true",
help="Merge individual input fastq's into left/right/single" + \
" file(s) [default: %default]")
p.set_trinity_opts()
p.set_grid()
opts, args = p.parse_args(args)
if len(args) not in (1, 2):
sys.exit(not p.print_help())
inparam, = args[:1]
assert op.exists(inparam)
genome = args[1] if len(args) == 2 else None
method = "GG" if genome is not None else "DN"
paired = opts.paired
merge = opts.merge
thome = opts.trinity_home
use_bam = opts.use_bam
gg_cpu = opts.gg_cpu
pf = inparam.split(".")[0]
tfolder = "{0}_{1}".format(pf, method)
cwd = os.getcwd()
mkdir(tfolder)
os.chdir(tfolder)
flist = iglob("../" + inparam, opts.names)
if paired:
f1 = [x for x in flist if "_1_" in x or ".1." in x or "_1." in x]
f2 = [x for x in flist if "_2_" in x or ".2." in x or "_2." in x]
assert len(f1) == len(f2)
if merge:
r1, r2 = "left.fastq", "right.fastq"
reads = ((f1, r1), (f2, r2))
else:
if merge:
r = "single.fastq"
reads = ((flist, r), )
if merge:
for fl, r in reads:
fm = FileMerger(fl, r)
fm.merge(checkexists=True)
cmd = op.join(thome, "Trinity")
cmd += " --seqType fq --max_memory {0} --CPU {1}".format(
opts.max_memory, opts.cpus)
cmd += " --min_contig_length {0}".format(opts.min_contig_length)
if opts.bflyGCThreads:
cmd += " --bflyGCThreads {0}".format(opts.bflyGCThreads)
if method == "GG":
cmd += " --genome {0} --genome_guided_max_intron {1}".format(
genome, opts.max_intron)
if use_bam:
cmd += " --genome_guided_use_bam {0}".format(use_bam)
if gg_cpu:
cmd += " --genome_guided_CPU {0}".format(gg_cpu)
if opts.grid and opts.grid_conf_file:
cmd += " --grid_conf_file={0}".format(opts.grid_conf_file)
if paired:
if merge:
cmd += " --left {0} --right {1}".format(reads[0][-1], reads[1][-1])
else:
for lf, rf in zip(f1, f2):
cmd += " --left {0}".format(lf)
cmd += " --right {0}".format(rf)
else:
if merge:
cmd += " --single {0}".format(reads[0][-1])
else:
for f in flist:
cmd += " --single {0}".format(f)
if opts.extra:
cmd += " {0}".format(opts.extra)
cmd += " --bypass_java_version_check"
runfile = "run.sh"
write_file(runfile, cmd)
os.chdir(cwd)
def compare(args):
"""
%prog compare pasa_db_name genome.fasta transcripts.fasta [annotation.gff]
Run the PASA annotation comparison pipeline
If annotation.gff file is provided, the PASA database is loaded with the annotations
first before starting annotation comparison. Otherwise, it uses previously
loaded annotation data.
Using the `--prepare` option creates a shell script with the run commands without
executing the pipeline
"""
p = OptionParser(compare.__doc__)
p.set_pasa_opts(action="compare")
p.add_option(
"--prepare",
default=False,
action="store_true",
help="Prepare PASA run script with commands [default: %default]")
p.set_grid()
p.set_grid_opts()
opts, args = p.parse_args(args)
if len(args) not in (3, 4):
sys.exit(not p.print_help())
pasa_db, genome, transcripts, = args[:3]
annotation = args[3] if len(args) == 4 else None
PASA_HOME = opts.pasa_home
if not op.isdir(PASA_HOME):
logging.error(
"PASA_HOME={0} directory does not exist".format(PASA_HOME))
sys.exit()
launch_pasa = which(op.join(PASA_HOME, "scripts", \
"Launch_PASA_pipeline.pl"))
grid = opts.grid
prepare, runfile = opts.prepare, "run.sh"
os.chdir(pasa_db)
if prepare:
write_file(runfile, "") # initialize run script
if opts.grid and not opts.threaded:
opts.threaded = opts.cpus
acfw = must_open(acconf, "w")
print >> acfw, annotCompare_conf.format("{0}_pasa".format(pasa_db), \
opts.pctovl, opts.pct_coding, opts.pctid_prot, opts.pctlen_FL, \
opts.pctlen_nonFL, opts.orf_size, opts.pct_aln, opts.pctovl_gene, \
opts.stompovl, opts.trust_FL, opts.utr_exons)
acfw.close()
if op.exists("{0}.clean".format(transcripts)):
transcripts = "{0}.clean".format(transcripts)
accmd = "{0} -c {1} -A -g {2} -t {3} --GENETIC_CODE {4}".format(launch_pasa, \
acconf, genome, transcripts, opts.genetic_code)
if annotation:
accmd += " -L --annots_gff3 {0}".format(annotation)
if prepare:
write_file(runfile, accmd, append=True)
else:
sh(accmd, grid=grid, grid_opts=opts)