def random_selecet(a):
fastq, number = a
length = int(
os.pipe(
"zcat {} | parallel --pipe wc -l | awk '{i+=$1}END{print i}'".
format(fastq)))
fastq_handle = SeqIO(gzip.open(fastq), 'fastq')
for record in fastq_handle:
num = random.randint(length)
if num < number:
print(record)
break
# coding: utf-8
from Bio import SeqIO
peptides = SeqIO('Data/anti_microbial_peptide.fasta', 'fasta')
peptides = SeqIO.parse('Data/anti_microbial_peptide.fasta', 'fasta')
peptides
peptides = []
for seq in SeqIO.p('Data/anti_microbial_peptide.fasta', 'fasta'):
peptides.append(seq)
for seq in SeqIO.parse('Data/anti_microbial_peptide.fasta', 'fasta'):
peptides.append(seq)
peptides
import pandas as pd
from protein_sequence_features import protein_features
from protein_sequence_features import protein_features
features = map(protein_features, peptides)
features[0]
peptides_dataset = pd.DataFrame(features, index=map(lambda x: x.id, peptides))
peptides_dataset
peptides_dataset.to_csv('Data/anti_microbial_peptide_features.csv')
from sklearn.manifold import TSNE
tsne = TSNE()
tsne.fit(peptides_dataset.values)
X = tsne.fit_transform(peptides_dataset.values)
import matplotlib.pyplot as plt
import seaborn as sns
plt.scatter(X)
plt.scatter(X[:, 0], X[:, 1])
plt.imshow()
plt.show()
def sss(par):
subsample = bool(par['subsample'])
select = bool(par['select'])
randomize = bool(par['randomize'])
if bool(par['out_f']):
n = par['split']
#openw = bz2.BZ2File if par['out_f'].endswith(".bz2") else open
if n == 1:
out_stream = [utils.openw(par['out_f'])]
else:
out_stream = [
utils.openw(par['out_f'] + str(r).zfill(len(str(n))) + ".fna" +
(".bz2" if par['out_f'].endswith(".bz2") else ""))
for r in range(n)
]
else:
out_stream = [sys.stdout] # larger buffer?
if select:
if os.path.exists(par['ids']):
#openr = bz2.BZ2File if par['ids'].endswith(".bz2") else open
es = [s.strip().split('\t')[0] for s in utils.openr(par['ids'])]
else:
es = [(s.split("$")[1] if s.count("$") else s)
for s in par['ids'].split(":::")]
es = set(es)
all_reads = []
nstreams = len(out_stream)
p = par['subsample']
#reads = reader( par['inp_f'], par['min_len'], par['max_len'] )
cind = 0
lmin, lmax = par['min_len'], par['max_len']
for r in SeqIO.parse(utils.openr(par['inp_f']), "fasta"):
if lmin and len(r.seq) < lmin:
continue
if lmax and len(r.seq) > lmax:
continue
if select:
if par['reverse']:
if r.id in es:
continue
elif r.id not in es:
continue
if subsample and rnd.random() > p:
continue
if randomize:
all_reads.append(r)
continue
SeqIO.write(r, out_stream[cind], "fasta")
cind = (cind + 1) % nstreams
"""
for r in reads:
if select and r.n not in es:
continue
if subsample and rnd.random() > p:
continue
if randomize:
all_reads.append( r )
continue
out_stream[cind].write( str(r) )
cind = (cind + 1) % nstreams
"""
if randomize:
rnd.shuffle(all_reads)
step = len(all_reads) / nstreams
for i, r in enumerate(all_reads):
#out_stream[cind].write( str(r) )
SeqIO(r, out_stream[cind], "fasta")
if not i % step:
cind = (cind + 1) % nstreams
for o in out_stream:
o.close()
