def _aln(ref, fastq, tmp="/tmp", threads=8, threshold=0.05):
sai = os.path.join(tmp, '%09d.sai' % random.randrange(0, 1e10))
with file_transaction(sai) as tx:
cmd = ("bwa aln -n {threshold} -t {threads} "
"{ref} {fastq} > {tx}").format(**locals())
run(cmd)
return sai
def samse_aln(ref, reads, bamout, tmp="/tmp", threads=8, threshold=0.05):
with bwa_index(ref) as bwaidx:
r_sai = _aln(bwaidx, reads, tmp, threads, threshold)
samse = ("bwa samse {ref} {r_sai} {reads} | samtools view -bSF0x0004 - "
"| samtools sort -f -m 8 - {bam_sorted}")
with file_transaction(bam_sorted) as tx:
run(samse)
return bam_sorted
def index_bam(bam_file):
"""
Build an index for a bam file.
parameters
bam_file : alignment file path
returns
index file name : string
"""
bam_index = bam_file + '.bai'
if not file_exists(bam_index):
with file_transaction(bam_index) as tx_out_file:
run('samtools index %s %s' % (bam_file, tx_out_file))
return bam_index
def sampe_aln(ref, reads, bam_sorted, tmp="/tmp", threads=1, threshold=0.05):
r1, r2 = tmp_split_reads(reads, tmp)
with bwa_index(ref) as bwaidx:
r1_sai = _aln(bwaidx, r1, tmp, threads, threshold)
r2_sai = _aln(bwaidx, r2, tmp, threads, threshold)
sampe = ("bwa sampe {ref} {r1_sai} {r2_sai} {r1} {r2} "
"| samtools view -bSF0x0004 - "
"| samtools sort -f -m 8 - {bam_sorted}").format(ref=bwaidx,
r1_sai=r1_sai,
r2_sai=r2_sai,
r1=r1,
r2=r2,
bam_sorted=bam_sorted)
with file_transaction(bam_sorted) as tx:
run(sampe)
return bam_sorted
def extract_fastq(bam, out_fastq):
''' Uses bedtools bamtofastq function to extract reads from bam
Args:
bam (string): path to bam alignment file
out_fastq (string): output fastq to write to
Returns:
out_fastq (string): path to written output
>> bam = 'Tara_test1_vs_Simons_LoCos_Conc.pctid95.overlap0.minlen100.bam'
>> outfastq = 'testout.fastq'
>> extract_fastq(bam, outfastq) == out_fastq
'''
with file_transaction(out_fastq) as temp_oh:
cmd = "bedtools bamtofastq -i {bam} -fq {fastq}".format(bam=bam,
fastq=temp_oh)
run(cmd)
return out_fastq
def sag_checkm_completeness(fasta, cores):
'''run checkm lineage_wf on SAG to get completeness values
Args:
fasta (str): full path to SAG genomic contigs in fasta format
cores (int): number of cores to use to run checkm
Returns:
"completeness" statistics as a pandas dataframe
'''
logger.info("Running checkm on %s " % fasta)
fasta = op.abspath(fasta)
if op.isdir == True or op.exists == False:
return None
with tmp_dir() as tdir:
bindir = op.join(tdir, "bindir")
safe_makedir(bindir)
outdir = op.join(tdir, "outdir")
safe_makedir(outdir)
tmp_fasta = op.join(bindir, op.basename(fasta))
try:
shutil.copy(fasta, tmp_fasta)
assert op.exists(tmp_fasta)
print(tmp_fasta, "created")
except Exception, e:
print("copying %s to the temporary directory failed, %s" % (fasta, e))
return None
logger.info("Running lineage workflow on %s" % fasta)
completeness_tsv = op.join(outdir, "completeness.tsv")
cmd = "checkm lineage_wf -f {outfile} --tab_table -q -x fasta -t {cores} {binpath} {outdir}".format(outfile=completeness_tsv, outdir=outdir, cores=cores, binpath=bindir)
logger.info("running checkm lineage, command is: {cmd}".format(**locals()))
run(cmd)
completeness = pd.read_csv(completeness_tsv, sep="\t", header=0)
def run_seqtk_sample(fastq, outfile, n, seed=37):
"""Subsample incoming paired-end fastqs to `n` reads (serially).
Args:
fastqs (str): path to fastq
outfile (str): path of output fastq paths; output files are always gzipped
n (int): number of subsampled reads
seed (int): for random selection of reads
Returns:
str: subsampled reads file path
"""
if file_exists(outfile):
return outfile
logger.info("Subsampling to %d reads" % n)
with file_transaction(outfile) as tx:
cmd = "seqtk sample -s {seed} {fastq} {number} | gzip > {out}".format(
seed=seed, fastq=fastq, number=n, out=tx)
run(cmd)
print("%s created" % outfile)
return outfile
def bwa_mem(fastq, out_file, reference, options, cores=1):
"""
align reads using bwa mem.
parameters
fastq : path to reads
out_file : path to aligned reads bam
index : path to bwa index
options : bwa mem options
cores : int
returns
output file path : string
"""
if file_exists(out_file):
return out_file
predefined_options = [('-t', False)]
if options is not None:
options = filter_options(options, predefined_options)
opts = " ".join(options)
else:
opts = ""
logger.info("Mapping %s to %s using bwa mem" % (fastq, reference))
reference = bwa_index(reference)
with file_transaction(out_file) as tx_out_file:
cmd = ("bwa mem -t {cores} {options} {index} {fastq} | samtools view "
"-ShuF4q2 - | samtools sort -o -m 8G - tmp > {result}"
).format(cores=cores,
options=opts,
index=reference,
fastq=fastq,
result=tx_out_file)
run(cmd)
index_bam(tx_out_file)
return out_file