def pygmes(self, fasta, db):
outdir = os.path.join(self.cfg["outdir"], "workfiles", "pygmes")
faafile = os.path.join(outdir, "predicted_proteins.faa")
bedfile = os.path.join(outdir, "predicted_proteins.bed")
self._clean_fasta = os.path.join(
outdir, "gmesclean_{}".format(os.path.basename(fasta)))
from pygmes import pygmes
# check if we need to launch
need_run = False
if not file().exists(faafile) or not file().exists(bedfile):
need_run = True
elif file.isnewer(fasta, faafile):
need_run = True
if need_run:
pygmes(fasta, outdir, db=db, clean=True, ncores=self.cfg["ncores"])
# check if pg worked
if os.path.exists(faafile) and os.path.exists(bedfile):
if os.stat(faafile).st_size == 0 or os.stat(bedfile).st_size == 0:
logging.warning("No predicted proteins")
self.write_outfile()
exit(1)
else:
return (faafile, bedfile)
else:
logging.warning("No predicted proteins, pyfaidx failed")
self.write_outfile()
exit(1)
def runPlacedHMM(self, hmmfile, proteinfaa, bedfile):
# run hmmer and strip down
# define output files
hmmDir = os.path.join(self.cfg["outdir"], "workfiles", "hmmer",
"estimations")
file.isdir(hmmDir)
hmmOut = os.path.join(hmmDir, "placement.tsv")
hmmOus = os.path.join(hmmDir, "placement.out")
hitOut = os.path.join(hmmDir, "hits.tsv")
h = hmmer(
"hmmsearch",
proteinfaa,
hmmOut,
self.cfg["debug"],
touch=self.cfg["touch"],
)
if h.doIneedTorun(
self.cfg["force"]) or self.cfg["fplace"] or file.isnewer(
hmmfile, hmmOut):
logging.info("Running hmmer for chosen locations")
h.run(
hmmOus,
hmmfiles=hmmfile,
modus=self.cfg["dbinfo"]["modus"],
evalue=self.cfg["evalue"],
cores=self.cfg["ncores"],
training=self.cfg["training"],
)
# clean hmmer outpout
logging.info("Processing Hmmer results")
hitOut = h.clean(hmmOut, bedfile, hitOut, self.cfg["mindist"])
return hitOut
def gmes(self, fasta):
"""
predict proteins using gmes
"""
logging.debug("Starting gmes function")
gmesDir = os.path.join(self.cfg["outdir"], "workfiles", "gmes")
file.isdir(gmesDir)
gmesOut = os.path.join(gmesDir, "prot_seq.faa")
gtffile = os.path.join(gmesDir, "genemark.gtf")
inputfasta = os.path.abspath(os.path.join(gmesDir, "input.fna"))
# GeneMark-ES
g = gmes("runGMES", fasta, [gtffile, gmesOut], touch=self.cfg["touch"])
logging.debug("Defined gmes run")
if g.doIneedTorun(self.cfg["force"]):
# rename fasta entries, so we dont have white spaces in them
# can be turned of via cleanfasta in config file
if not self.cfg["touch"]:
g.input = base.clearFastaNames(fasta, inputfasta)
else:
g.input = inputfasta
logging.info("Running GeneMark-ES")
g.run(cores=self.cfg["ncores"])
else:
logging.debug("I do not need to run gmes, output exists:")
logging.debug(gtffile)
# always check if gtffile exists, if not Genemark-ES failed and
# we can stop here
if not file.exists(gtffile):
# log and document failing
# then stop pipeline
logging.error("GeneMark-ES failed on this bin")
self.write_outfile()
exit(1)
elif self.cfg["clean"]:
# clean temp dirs
_tmpdirs = ["data", "run", "info", "output/data", "output/gmhmm"]
tempdirs = [os.path.join(gmesDir, x) for x in _tmpdirs]
g.cleanup(tempdirs)
# make a bed file from GTF
bedf = os.path.join(gmesDir, "proteins.bed")
if self.cfg["force"] or file.isnewer(gtffile,
bedf) and not self.cfg["touch"]:
logging.info("Extracting protein locations")
bedf = base.gmesBED(gtffile, bedf)
# touch files expected for next step
if self.cfg["touch"]:
g.touch([bedf, gmesOut])
self._clean_fasta = inputfasta
return (gmesOut, bedf)
def doIneedTorun(self, force=False):
logging.debug("Testing if I need to run this step")
if force or self.touchonly:
return True
else:
for p in self.output_test:
x = file.isnewer(self.input, p)
if x:
logging.debug(f"Need to run because of file: {p}")
return x
return x