def get_read_limit(fasta, readLimit, verbose):
"""Return read limit and libraries."""
# limit no. of reads to align as fraction of genome size
limit = 0
if readLimit:
stats = fasta_stats(open(fasta))
fastaSize = int(stats.split("\t")[2])
limit = int(readLimit * fastaSize)
if verbose:
sys.stderr.write(" Aligning %s mates per library...\n" % limit)
return limit
def run_scaffolding(outdir, scaffoldsFname, fastq, libraries, reducedFname, mapq, threads, \
joins, limit, iters, sspacebin, verbose, \
identity, overlap, minLength, lib=""):
"""Execute scaffolding step."""
# run scaffolding using libraries with increasing insert size in multiple iterations
pout = reducedFname
i = 0
#for i, (libnames, libFs, libRs, orients, libIS, libISStDev) in enumerate(libraries, 1):
while i < len(libraries):
libnames, libFs, libRs, orients, libIS, libISStDev = libraries[i]
i += 1
for j in range(1, iters+1):
if verbose:
sys.stderr.write(" iteration %s.%s ...\n"%(i,j))
out = os.path.join(outdir, "_sspace.%s.%s"%(i, j))
lib = ""
# run fastq scaffolding
fastq2sspace(out, open(pout), lib, libnames, libFs, libRs, orients, \
libIS, libISStDev, threads, mapq, limit, joins, \
sspacebin, verbose=0)
# store out info
pout = out+".fa"
# link output ie out/_sspace.1.1/_sspace.1.1.scaffolds.fasta --> out/_sspace.1.1.scaffolds.fasta
targetout = os.path.join(os.path.basename(out), os.path.basename(out+".final.scaffolds.fasta"))
symlink(targetout, pout)
# if number of gaps larger than 1%, run gap closer & reduction
stats = fasta_stats(open(pout))
fastaSize = int(stats.split('\t')[2])
gapSize = int(stats.split('\t')[-2])
if 1.0 * gapSize / fastaSize > 0.01:
# close gaps
if verbose:
sys.stderr.write(" closing gaps ...\n")
nogapsFname = ".".join(pout.split(".")[:-1]) + ".filled.fa"
basename = "_sspace.%s.%s._gapcloser"%(i, j)
run_gapclosing(outdir, mapq, [libraries[i-1],], nogapsFname, pout, \
threads, limit, 1, 0, basename)
pout = nogapsFname
# reduce
'''reducedFname = ".".join(pout.split(".")[:-1]) + ".reduced.fa"
with open(reducedFname, "w") as out:
fasta2homozygous(out, open(nogapsFname), identity, overlap, \
minLength, libraries, limit, threads)
# update pout
pout = reducedFname #nogapsFname'''
# update library insert size estimation, especially for mate-pairs
libraries = get_libraries(fastq, pout, mapq, threads, verbose=0)
# create symlink to final scaffolds or pout
symlink(pout, scaffoldsFname)
return libraries
def redundants(
fastq,
fasta,
outdir,
mapq,
threads,
identity,
overlap,
minLength,
joins,
readLimit,
iters,
sspacebin,
reduction=1,
scaffolding=1,
gapclosing=1,
cleaning=1,
verbose=1,
log=sys.stderr,
):
"""Launch redundans pipeline."""
# redirect stderr
# sys.stderr = log
# prepare outdir or quit if exists
if os.path.isdir(outdir):
sys.stderr.write("Directory %s exists!\n" % outdir)
sys.exit(1)
else:
os.makedirs(outdir)
# REDUCTION
contigsFname = os.path.join(outdir, "contigs.fa")
reducedFname = os.path.join(outdir, "contigs.reduced.fa")
# link contigs & genome
symlink(fasta, contigsFname)
# get read limit & libraries
limit = get_read_limit(contigsFname, readLimit, verbose)
libraries = get_libraries(fastq, contigsFname, mapq, threads, verbose)
if reduction:
if verbose:
sys.stderr.write("%sReduction...\n" % timestamp())
sys.stderr.write(
"#file name\tgenome size\tcontigs\theterozygous size\t[%]\theterozygous contigs\t[%]\tidentity [%]\tpossible joins\thomozygous size\t[%]\thomozygous contigs\t[%]\n"
)
with open(reducedFname, "w") as out:
info = fasta2homozygous(out, open(contigsFname), identity, overlap, minLength, libraries, limit, threads)
else:
symlink(contigsFname, reducedFname)
# update fasta list
fastas = [contigsFname, reducedFname]
# update read limit using reduced assembly as reference
limit = get_read_limit(reducedFname, readLimit, verbose)
# SCAFFOLDING
scaffoldsFname = os.path.join(outdir, "scaffolds.fa")
if scaffolding:
if verbose:
sys.stderr.write("%sScaffolding...\n" % timestamp())
# estimate read limit
libraries = run_scaffolding(
outdir,
scaffoldsFname,
fastq,
libraries,
reducedFname,
mapq,
threads,
joins,
limit,
iters,
sspacebin,
verbose,
identity,
overlap,
minLength,
)
else:
symlink(reducedFname, scaffoldsFname)
# update fasta list
fastas += sorted(glob.glob(os.path.join(outdir, "_sspace.*.fa")))
fastas.append(scaffoldsFname)
# GAP CLOSING
## gapclosing is only necessary after scaffolding
nogapsFname = os.path.join(outdir, "scaffolds.filled.fa")
if gapclosing and libraries:
if verbose:
sys.stderr.write("%sGap closing...\n" % timestamp())
run_gapclosing(outdir, mapq, libraries, nogapsFname, scaffoldsFname, threads, limit, iters, verbose)
else:
symlink(scaffoldsFname, nogapsFname)
# update fasta list
fastas += sorted(glob.glob(os.path.join(outdir, "_gap*.fa")))
fastas.append(nogapsFname)
# FASTA STATS
if verbose:
sys.stderr.write("%sReporting statistics...\n" % timestamp())
# report stats
sys.stderr.write("#fname\tcontigs\tbases\tGC [%]\tcontigs >1kb\tbases in contigs >1kb\tN50\tN90\tNs\tlongest\n")
for fn in fastas:
sys.stderr.write(fasta_stats(open(fn)))
# Clean-up
# rm fq.is.txt
if cleaning:
if verbose:
sys.stderr.write("%sCleaning-up...\n" % timestamp())
for root, dirs, fnames in os.walk(outdir):
for fn in filter(lambda x: not x.endswith((".fa", ".fasta", ".stats")), fnames):
os.unlink(os.path.join(root, fn))