def get_nucleotide(args):
"""Function to obain sequence content around sites"""
if not args.out:
args.out = '.'.join(os.path.basename(args.bed).split('.')[0:-1])
chunks = ChunkList.read(args.bed, strand_col=args.strand)
params = _NucleotideParameters(args.up, args.down, args.fasta,
args.dinucleotide)
sets = chunks.split(bases=10000)
pool = Pool(processes=args.cores)
tmp = pool.map(_nucleotideHelper, list(zip(sets,
itertools.repeat(params))))
pool.close()
pool.join()
result = np.zeros(params.matsize)
n = 0.0
for i in tmp:
result += i[0]
n += i[1]
result = result // n
if args.norm:
normfreqs = seq.getNucFreqs(params.fasta, params.nucleotides)
result = result / np.reshape(np.repeat(normfreqs, result.shape[1]),
result.shape)
#save text output
out = np.hstack(
(np.array(params.nucleotides)[:, np.newaxis], result.astype('|S8')))
np.savetxt(args.out + '.nucfreq.txt', out, delimiter="\t",
fmt='%s') #,fmt = '%1.4g')
def get_pwm(args, bases=50000, splitsize=1000):
"""Functiono obtain PWM around ATAC insertion"""
if not args.out:
args.out = '.'.join(os.path.basename(args.bam).split('.')[0:-1])
chrs = read_chrom_sizes_from_fasta(args.fasta)
if args.bed is None:
chunks = ChunkList.convertChromSizes(chrs,
splitsize=splitsize,
offset=args.flank)
sets = chunks.split(items=bases / splitsize)
else:
chunks = ChunkList.read(args.bed,
chromDict=chrs,
min_offset=args.flank)
sets = chunks.split(bases=bases)
params = _PWMParameters(bam=args.bam,
up=args.flank,
down=args.flank,
fasta=args.fasta,
lower=args.lower,
upper=args.upper,
atac=args.atac,
sym=args.sym)
pool = Pool(processes=args.cores)
tmp = pool.map(_pwmHelper, zip(sets, itertools.repeat(params)))
pool.close()
pool.join()
n = 0.0
result = np.zeros((len(params.nucleotides), params.up + params.down + 1))
for i in tmp:
result += i[0]
n += i[1]
result /= n
if args.bed:
normfreqs = seq.getNucFreqsFromChunkList(chunks, args.fasta,
params.nucleotides)
else:
normfreqs = seq.getNucFreqs(args.fasta, params.nucleotides)
result = result / np.reshape(np.repeat(normfreqs, result.shape[1]),
result.shape)
if args.sym:
#Symmetrize
left = result[:, 0:(args.flank + 1)]
right = result[:, args.flank:]
rightflipped = np.fliplr(np.flipud(right))
combined = (left + rightflipped) / 2
result = np.hstack(
(combined, np.fliplr(np.flipud(combined[:, 0:args.flank]))))
#save
pwm = PWM(result, args.flank, args.flank, params.nucleotides)
pwm.save(args.out + '.PWM.txt')
def get_pwm(args, bases = 50000, splitsize = 1000):
"""Functiono obtain PWM around ATAC insertion"""
if not args.out:
args.out = '.'.join(os.path.basename(args.bam).split('.')[0:-1])
chrs = read_chrom_sizes_from_fasta(args.fasta)
if args.bed is None:
chunks = ChunkList.convertChromSizes(chrs, splitsize = splitsize, offset = args.flank)
sets = chunks.split(items = bases/splitsize)
else:
chunks = ChunkList.read(args.bed, chromDict = chrs, min_offset = args.flank)
sets = chunks.split(bases = bases)
params = _PWMParameters(bam = args.bam, up = args.flank, down = args.flank, fasta = args.fasta,
lower = args.lower, upper = args.upper, atac = args.atac, sym = args.sym)
pool = Pool(processes = args.cores)
tmp = pool.map(_pwmHelper, zip(sets,itertools.repeat(params)))
pool.close()
pool.join()
n = 0.0
result = np.zeros((len(params.nucleotides), params.up + params.down + 1))
for i in tmp:
result += i[0]
n += i[1]
result /= n
if args.bed:
normfreqs = seq.getNucFreqsFromChunkList(chunks, args.fasta, params.nucleotides)
else:
normfreqs = seq.getNucFreqs(args.fasta, params.nucleotides)
result = result / np.reshape(np.repeat(normfreqs,result.shape[1]),result.shape)
if args.sym:
#Symmetrize
left = result[:,0:(args.flank + 1)]
right = result[:,args.flank:]
rightflipped = np.fliplr(np.flipud(right))
combined = (left + rightflipped) / 2
result = np.hstack((combined, np.fliplr(np.flipud(combined[:,0:args.flank]))))
#save
pwm = PWM(result, args.flank, args.flank, params.nucleotides)
pwm.save(args.out + '.PWM.txt')
def get_nucleotide(args):
"""Function to obain sequence content around sites"""
if not args.out:
args.out = '.'.join(os.path.basename(args.bed).split('.')[0:-1])
chunks = ChunkList.read(args.bed, strand_col = args.strand)
params = _NucleotideParameters(args.up, args.down, args.fasta, args.dinucleotide)
sets = chunks.split(bases = 10000)
pool = Pool(processes = args.cores)
tmp = pool.map(_nucleotideHelper, zip(sets,itertools.repeat(params)))
pool.close()
pool.join()
result = np.zeros(params.matsize)
n = 0.0
for i in tmp:
result += i[0]
n += i[1]
result = result / n
if args.norm:
normfreqs = seq.getNucFreqs(params.fasta, params.nucleotides)
result = result / np.reshape(np.repeat(normfreqs,result.shape[1]),result.shape)
#save text output
out = np.hstack((np.array(params.nucleotides)[:,np.newaxis], result.astype('|S8')))
np.savetxt(args.out+'.nucfreq.txt', out, delimiter="\t", fmt = '%s')#,fmt = '%1.4g')
