def build_index(fa, chrom, site, strand, rlen, thread, out_dir, seq, seq_flag):
print('Build index...')
if strand == '+':
start = site - (rlen - 10)
end = site + (rlen - 20)
offset = rlen - 10
else:
start = site - (rlen - 20)
end = site + (rlen - 10)
offset = rlen - 20
index_path = os.path.join(out_dir, 'sgRNA.fa')
if seq_flag:
os.symlink(seq, index_path)
else:
# fetch sgRNA region sequence
with open(index_path, 'w') as out:
out.write('>sgRNA_region\n')
out.write(
dna_to_rna(fa.fetch(chrom, start, end), strand=strand) + '\n')
# build index
if which('bowtie2-build'):
command = 'bowtie2-build -q --threads %s %s %s'
command = command % (thread, index_path, index_path)
run_command(command, 'Error: cannot build index for sgRNA!')
else:
sys.exit('Error: no bowtie2-build installed!')
return index_path, offset
def run_fseq(folder, bed, strand, flength, wig_flag, percent):
prefix = os.path.splitext(bed)[0]
# create bed files
temp_dir = tempfile.mkdtemp()
bed_f = os.path.join(folder, bed)
# run fseq
command = 'fseq -f 0 -l %s -of bed -o %s %s' % (flength, temp_dir, bed_f)
run_command(command, 'Error in F-seq!')
# cat fseq files
peak_f = os.path.join(folder, prefix + '_fseq.bed')
cat_files(temp_dir, peak_f)
# resize peaks
resized_peak_f = os.path.join(folder, prefix + '_peak.bed')
resize_peak(peak_f, bed_f, resized_peak_f, strand, percent)
# create wig files
if wig_flag:
# run fseq
temp_dir = tempfile.mkdtemp()
command = 'fseq -f 0 -l %s -o %s %s' % (flength, temp_dir, bed_f)
run_command(command, 'Error in F-seq!')
# cat fseq wig files
wig = prefix + '_fseq.wig'
wig_f = os.path.join(folder, wig)
cat_files(temp_dir, wig_f, is_wig=True, name=wig)
return resized_peak_f
def bg_to_bw(bg, chrom_size):
if not os.path.isfile(chrom_size):
sys.exit('No chrom size file: %s!' % chrom_size)
prefix = os.path.splitext(os.path.split(bg)[-1])[0]
bw = prefix + '.bw'
command = 'bedGraphToBigWig %s %s %s' % (bg, chrom_size, bw)
run_command(command, 'Could not convert bg to bw!')
return bw
def run_stringtie(bam, gtf, dir_path, thread):
if not os.path.isfile(bam):
sys.exit('No BAM file: %s!' % bam)
if not os.path.isfile(gtf):
sys.exit('No GTF file: %s' % gtf)
fname = os.path.basename(bam)
prefix = os.path.splitext(fname)[0]
outf = os.path.join(dir_path, prefix + '_stringtie.gtf')
command = 'stringtie -G %s -p %s -o %s --rf %s' % (gtf, thread, outf, bam)
run_command(command, 'Error in StringTie!')
return outf
def bowtie2_align(index, read, thread, out_dir):
print('Align reads...')
if which('bowtie2'):
bam = os.path.join(out_dir, 'cs.bam')
sam = tempfile.NamedTemporaryFile('w+')
command = 'bowtie2 --quiet --end-to-end -p %s -x %s -U %s -S %s'
command = command % (thread, index, read, sam.name)
run_command(command, 'Error in bowtie2 alignment!')
sam.seek(0)
with pysam.AlignmentFile(sam.name, 'r') as sam_f:
with pysam.AlignmentFile(bam, 'wb', template=sam_f) as bam_f:
for read in sam_f:
if not read.is_unmapped:
bam_f.write(read)
sam.close()
return bam
else:
sys.exit('Error: no bowtie2 installed!')
def merge_stringtie(gtf_list, gtf, dir_path, prefix, thread):
outf = os.path.join(dir_path, prefix + '.gtf')
command = 'stringtie --merge -G %s -p %s -o %s %s' % (gtf, thread, outf,
'\t'.join(gtf_list))
run_command(command, 'Error in StringTie --merge!')
return outf