def generate_bedfile(ref_fasta, kmers, output_bedfile, tmpdir='bedgentmp', threads='2', logfile=None):
"""
Given a reference FASTA file and a fasta-formatted set of kmers, will generate a coverage bedfile for the reference
FASTA by mapping the kmers back to the FASTA.
:param ref_fasta: Path to reference FASTA.
:param kmers: Path to FASTA-formatted kmer file.
:param output_bedfile: Path to output bedfile.
:param tmpdir: Temporary directory to store intermediate files. Will be deleted upon method completion.
:param threads: Number of threads to use for analysis. Must be a string.
"""
if not os.path.isdir(tmpdir):
os.makedirs(tmpdir)
# First, need to generate a bam file - align the kmers to a reference fasta genome.
bbtools.bbmap(ref_fasta, kmers, os.path.join(tmpdir, 'out.bam'), threads=threads, ambig='best',
perfectmode='true')
# Once the bam file is generated, turn it into a sorted bamfile so that bedtools can work with it.
cmd = 'samtools sort {bamfile} -o {sorted_bamfile}'.format(bamfile=os.path.join(tmpdir, 'out.bam'),
sorted_bamfile=os.path.join(tmpdir, 'out_sorted.bam'))
if logfile:
with open(logfile, 'a+') as f:
f.write('Command: {}'.format(cmd))
subprocess.call(cmd, shell=True, stderr=f, stdout=f)
else:
subprocess.call(cmd, shell=True)
# Use bedtools to get genome coverage, so that we know what to mask.
cmd = 'bedtools genomecov -ibam {sorted_bamfile} -bga' \
' > {output_bed}'.format(sorted_bamfile=os.path.join(tmpdir, 'out_sorted.bam'),
output_bed=output_bedfile)
if logfile:
with open(logfile, 'a+') as f:
f.write('Command: {}'.format(cmd))
subprocess.call(cmd, shell=True, stderr=f, stdout=f)
else:
subprocess.call(cmd, shell=True)
shutil.rmtree(tmpdir)
def test_bbmap_command_paired():
out, err, cmd = bbtools.bbmap(reference='tests/dummy_fasta/test.fasta',
forward_in='tests/dummy_fastq/test_R1.fastq',
out_bam='tests/out.bam',
returncmd=True)
assert cmd == 'bbmap.sh ref=tests/dummy_fasta/test.fasta in=tests/dummy_fastq/test_R1.fastq' \
' in2=tests/dummy_fastq/test_R2.fastq out=tests/out.bam nodisk'
os.remove('tests/out.bam')
def test_bbmap_command_paired_kwargs():
out, err, cmd = bbtools.bbmap(reference='tests/dummy_fasta/test.fasta',
forward_in='tests/dummy_fastq/test_R1.fastq',
out_bam='tests/out.bam',
returncmd=True,
threads='3',
ordered='t')
assert 'bbmap.sh ref=tests/dummy_fasta/test.fasta in=tests/dummy_fastq/test_R1.fastq' \
' in2=tests/dummy_fastq/test_R2.fastq out=tests/out.bam nodisk' in cmd and 'ordered=t' in cmd and 'threads=3' in cmd
os.remove('tests/out.bam')
def create_bam(forward_reads, reverse_reads, reference_fasta, outdir):
# Map with bbmap default settings. This may or may not work.
# Getting some real weird results with this - try using bowtie2 and see what happens
# Bowtie2 gives really similar results with default settings. Will need to play with parameters lots tomorrow to
# try to figure out what's going on.
cmd = 'samtools faidx {reference_fasta}'.format(
reference_fasta=reference_fasta)
out, err = run_cmd(cmd)
write_to_logfile(logfile=os.path.join(outdir, 'log.txt'),
out=out,
err=err,
cmd=cmd)
bbtools.bbmap(
reference=reference_fasta,
forward_in=forward_reads,
reverse_in=reverse_reads,
out_bam=os.path.join(outdir, 'aligned.bam'),
subfilter=1
) # Limiting to one substitution allowed per read mapped might help - TBD
# Also sort and index the bamfile so pysam will be happy with us.
cmd = 'samtools sort {bamfile} -o {sorted_bamfile}'.format(
bamfile=os.path.join(outdir, 'aligned.bam'),
sorted_bamfile=os.path.join(outdir, 'aligned_sorted.bam'))
out, err = run_cmd(cmd)
write_to_logfile(logfile=os.path.join(outdir, 'log.txt'),
out=out,
err=err,
cmd=cmd)
cmd = 'samtools index {sorted_bamfile}'.format(
sorted_bamfile=os.path.join(outdir, 'aligned_sorted.bam'))
out, err = run_cmd(cmd)
write_to_logfile(logfile=os.path.join(outdir, 'log.txt'),
out=out,
err=err,
cmd=cmd)