contexts = []
split_length = 200
for fast5 in glob.glob(dirpath + '*.fast5'):
dataset = guppy_fast5_extraction(fast5)
for data in dataset:
read_id = data[0]
if read_id not in rDNA_read_ids:
continue
split_fastq = data[1].split()
read = split_fastq[6]
quality = split_fastq[8]
mod_base_table = data[2]
cpg_scores = make_methylation_summary(read, mod_base_table)
averaged_cpg = calculate_meth_stats(cpg_scores)
utilities.split_mapping_and_sam_analysis(split_length, read_id, read, quality, '../clive/rDNA_index/humRibosomal.fa')
lc = plot_read_structure('test', split_length, 0)
fig = plt.figure()
plt.subplots_adjust(left=0.2)
ax = fig.add_subplot()
x = []
y = []
for n, score in averaged_cpg.items():
x.append(n * 200)
y.append(score * 10000)
xa = []
ya = []
for n, base in enumerate(mod_base_table):
if base[1] != 0 and read[n] == 'A':
contexts.append(read[n-1:n+3])
xa.append(n)
ya.append(base[1] * 40)
def find_boundaries_from_fastq(fastq, split_length):
"""Find boundary containing reads from a .fastq file.
Args:
fastq (str): filename
split_length (split_length): split_length
Returns:
boundaries: list of (boundary coordinate, boundary rDNA coordinate,
direction, side of non rDNA, read, header)
"""
with open(fastq) as f:
boundaries = []
for n, each_fastq in enumerate(itertools.zip_longest(*[iter(f)] * 4)):
header = each_fastq[0].strip()
read = each_fastq[1].strip()
if len(read) < 40000:
continue
quality = each_fastq[3].strip()
make_temp_fastq(split_length, header, read, quality)
subprocess.run(
'bwa mem -M -x ont2d -t 5 '
'/home/yutaro/nanopore/clive/rDNA_index/'
'humRibosomal.fa temp_files/temp_fastq.fastq > '
'temp_files/temp_sam.sam',
shell=True,
stdout=FNULL,
stderr=subprocess.STDOUT)
# rDNA_coordinate=1 when using find_true_boundary2
temp_boundary = find_end_reads('temp_files/temp_sam.sam',
split_length,
rDNA_coordinate=1)
if temp_boundary:
header = header.split()[0]
boundary = int(temp_boundary[0])
direction = temp_boundary[1]
side = temp_boundary[2]
if direction == '+':
if side == 'right':
bound_seq = read[boundary:boundary + 10000]
with open('boundary_seq1.fa', 'a') as fw:
fw.write('>' + header + '\n')
fw.write(bound_seq + '\n\n')
else:
with open('boundary_seq2.fa', 'a') as fw:
fw.write('>' + header + '\n')
fw.write(bound_seq + '\n\n')
else:
bound_seq = read[boundary - 10000:boundary]
revcom = str(Seq(bound_seq).reverse_complement())
if side == 'left':
with open('boundary_seq1.fa', 'a') as fw:
fw.write('>' + header + '\n')
fw.write(revcom + '\n\n')
else:
with open('boundary_seq2.fa', 'a') as fw:
fw.write('>' + header + '\n')
fw.write(revcom + '\n\n')
plot_read_structure(header,
split_length,
savename='end_reads/' + header + '.png',
title=str(temp_boundary[0]))
continue
true_boundary = find_true_boundary2(header, read, quality,
temp_boundary)
if true_boundary:
boundaries.append(
(true_boundary[0], true_boundary[1], temp_boundary[1],
temp_boundary[2], read, header.split()[0]))
return boundaries
group2 = pd.read_pickle('group2.pkl')
num2id1 = {}
with open('boundary_seq1.fa') as f:
for n, line in enumerate(f):
if n % 3 == 0:
num2id1[n // 3] = line.strip()[1:]
num2id2 = {}
with open('boundary_seq2.fa') as f:
for n, line in enumerate(f):
if n % 3 == 0:
num2id2[n // 3] = line.strip()[1:]
for n, i in enumerate(group1):
if len(i) > 2:
for item in i:
plot_read_structure(num2id1[item],
200,
savename='boundary1/' + str(n) + '/' +
num2id1[item])
for n, i in enumerate(group2):
if len(i) > 2:
try:
os.mkdir('boundary2/' + str(n))
finally:
for item in i:
plot_read_structure(num2id2[item],
200,
savename='boundary2/' + str(n) + '/' +
num2id2[item])
quit()
seq_list1 = []
with open('boundary_seq1.fa') as f:
for fa in itertools.zip_longest(*[iter(f)] * 3):