def intron_from_line(line):
try:
if isinstance(line[3], str) and line[3] in '+-':
return Intron(line[0], line[1], line[2], line[3])
else:
return Intron(line[0], line[1], line[2])
except IndexError:
return Intron(line[0], line[1], line[2])
def _line_to_intron(line):
"""RegTools junctions to Intron."""
scaffold = line[0]
start = line[1] + int(line[-2].split(',')[0])
end = line[2] - int(line[-2].split(',')[1])
score = line[4]
return Intron(scaffold, start, end, support=score)
def extract_introns_from_gtf(file, file_out):
introns_p = []
unique = set()
for line in process_file(file):
if line[2] == 'transcript':
new_gene = True
gene = line[9]
elif line[2] == 'exon':
if new_gene:
new_gene = False
start = line[4]
else:
end = line[3] - 1
scaffold = line[0]
sign = line[6]
i = Intron(scaffold, start, end, gene=gene, strand=sign)
if ' '.join([scaffold, str(start), str(end)]) not in unique:
unique.add(' '.join([scaffold, str(start), str(end)]))
introns_p.append(i)
start = line[4]
with open(file_out, 'w') as f_out:
for intron in introns_p:
f_out.write('\t'.join([
str(x) for x in [
intron.scaffold, intron.start, intron.end, intron.gene,
intron.strand
]
]))
f_out.write('\n')
def intron_stats(file):
"""Calculate basic statistics of introns from a file."""
introns = []
for line in process_file(file):
i = Intron(line[0], line[1], line[2], support=line[3])
introns.append(i)
print('Number of introns:0', len(introns))
print('Mean support: ', sum([i.support for i in introns]) / len(introns))
print('Median support: ', median([intron.support for intron in introns]))
def choose_best_introns(file_in, file_out, cutoff):
"""
Choose one best intron over every position.
:param file_in: (str) Path to the .bed file with introns in format: scaffold start end support. All introns from
a scaffold must come one after another in the file, and within one scaffold introns have to be sorted by start.
:param file_out: (str) Path to the out file with best introns.
:param cutoff: (int) Minimum support of the best intron.
:return: Two dictionaries where key is scaffold and value is the list of introns on the scaffold:
one containing all the introns from the input file and one with the best introns.
"""
with open(file_out, 'w') as f_out:
best_introns = defaultdict(list)
all_introns = defaultdict(list)
chrom_old = 'scaffold_0'
start_old = 0
end_old = 0
score_old = 0
def write_junction():
junction = '\t'.join([str(x) for x in [chrom, start_old, end_old, score_old]])
f_out.write(junction)
f_out.write('\n')
best_introns[chrom].append(i)
for line in process_file(file_in):
chrom, start, end, score = line
i = Intron(chrom, start, end, support=score)
all_introns[chrom].append(i)
# only consider introns with high enough support
if score < cutoff:
continue
if chrom == chrom_old:
if start < end_old:
# still in the same intron
if score > score_old:
# one best intron in each position
start_old, end_old, score_old = start, end, score
else:
# in a new intron, so the old one has to be written down
if not start_old - end_old == 0:
write_junction()
start_old, end_old, score_old = start, end, score
else:
# new scaffold, so we need to write down the last intron
write_junction()
chrom_old, start_old, end_old, score_old = chrom, start, end, score
# now we need to write the last one
write_junction()
return all_introns, best_introns
def intron_dict(file):
my_introns = defaultdict(list)
for line in process_file(file):
intron = Intron(line[0],
line[1],
line[2],
gene=line[3],
strand=line[4])
my_introns[intron.scaffold].append(intron)
return my_introns
def file_to_seq_introns(file, margin):
"""Extract introns with sequences from a fasta file"""
introns = []
with open(file) as f_in:
for line in f_in.readlines():
if line[0] == '>':
line = re.split('[>:\-]', line.strip())
scaffold, start, end = line[1:]
else:
i = Intron(scaffold,
int(start),
int(end),
margin_left=margin,
margin_right=margin,
sequence=line.strip())
introns.append(i)
return introns