def test_Interval_sequence():
genome = pyfasta.Fasta('test/example.fa')
l1 = Interval.from_string('1:858-967:1', genome=genome)
l2 = Interval.from_string('1:858-967:-1', genome=genome)
print l1.sequence
print l2.sequence
assert l1.sequence != l2.sequence
def test_Interval_distance():
l1 = Interval.from_string('chr1:10858-10967:1')
l2 = Interval.from_string('chr1:10858-10967:-1')
assert l1.distance(l2) == 0
l3 = Interval.from_string('chr1:10968-10977:-')
print l1.distance(l3)
assert l1.distance(l3) == 1
assert l3.distance(l1) == 1
def test_Interval_contains():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-10964:-1')
l3 = Interval.from_string('chr1:20858-30001:-1')
l4 = Interval.from_string('chr1:30000-30003:-1')
assert l2 in l1
assert not l1 in l2
assert not l3 in l4
def test_Interval_overlaps():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-21000:-1')
l3 = Interval.from_string('chr1:20858-30000:-1')
l4 = Interval.from_string('chr1:30000-30001:-1')
assert l1.overlaps(l2)
assert l2.overlaps(l3)
assert not l1.overlaps(l3)
assert not l3.overlaps(l4)
def test_Interval_is_contiguous():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-21000:-1')
l3 = Interval.from_string('chr1:20858-30000:-1')
l4 = Interval.from_string('chr1:30000-30001:-1')
print l1.distance(l2) == 0
assert l1.is_contiguous(l2)
assert l2.is_contiguous(l3)
assert not l1.is_contiguous(l3)
assert l3.is_contiguous(l4)
def test_truncate():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = l1.truncate(100)
assert l2.start == 10000
assert l2.end == 10100
l1 = Interval.from_string('chr1:10000-10967:-1')
l2 = l1.truncate(100)
assert l2.start == 10867
assert l2.end == 10967
def test_Interval_merge():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-12964:-1')
l3 = Interval.from_string('chr1:10858-10964:-1')
merged = Interval.merge([l1,l2,l3])
assert len(merged) == 1
m = merged[0]
assert m.start == 10000
assert m.end == 12964
def intervals(self):
if self.bed_file is None:
return {}
if not self.bed_file.exists:
raise ValueError(f"can not find the bed file: {self.bed_file}")
intervals = defaultdict(list)
with open(self.bed_file) as fp:
for line in fp:
chrom, start, end, *_ = line.strip().split('\t')
intervals[chrom].append(
Interval(int(start), int(end), chrom=chrom))
return {chrom: Interval.merge(intervals[chrom]) for chrom in intervals}
def test_Interval_from_string():
a = Interval.from_string('chr1:10858-10967:1')
assert a.chrom == 'chr1'
assert a.start == 10858
assert a.end == 10967
assert a.strand == 1
a = Interval.from_string('chr1:10858-10967:-1')
assert a.chrom == 'chr1'
assert a.start == 10858
assert a.end == 10967
assert a.strand == -1
def test_Interval_span():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:20858-30000:-1')
l3 = Interval.from_string('chr2:20858-30000:-1')
ex = l1.span(l2)
print ex
assert ex.start == 10000
assert ex.end == 30000
ex = l2.span(l1)
assert ex.start == 10000
assert ex.end == 30000
try:
l1.span(l3)
assert False
except:
pass
def test_Interval_sub():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-10964:-1')
l3 = Interval.from_string('chr2:10858-10964:-1')
i = l1 - l2
print i
assert len(i)==2
l = i[0]
r = i[1]
assert l.start == 10000
assert l.end == 10858
assert r.start == 10964
assert r.end == 10967
assert len(l1 - l3) == 1
def load_amplicons(design, stats, opts):
amplicons = []
for row in csv.DictReader(file(design, 'U'), delimiter=opts.delimiter):
amp_loc = Interval.from_string(row[opts.amplicon_column])
trim_loc = Interval.from_string(row[opts.trim_column])
if not trim_loc in amp_loc:
print('trim location not contained in amplicon location, impossible trim', file=sys.stderr)
sys.exit(1)
amplicon = Amplicon(
chr=amp_loc.chrom, start=amp_loc.start, end=amp_loc.end, strand=amp_loc.strand,
trim_start = trim_loc.start, trim_end=trim_loc.end,
external_id = row[opts.id_column], stats = stats,
offset_allowed=opts.offset_allowed,
)
amplicons.append(amplicon)
# TODO: check that amplicons can be uniquely resolved
return amplicons
def within_interval(self, segment):
if not self.intervals:
return True
if segment.reference_name not in self.intervals:
return False
current_position = Interval(
segment.reference_start or 0,
segment.reference_end
or segment.reference_start + segment.query_length,
chrom=segment.reference_name)
for interval in self.intervals[segment.reference_name]:
if interval.distance(current_position) < self.flank_size:
return True
return False
def load_amplicons_from_header(header, stats, samfile, clip=True, load_pileups=True):
amplicons = []
for row in header['CO']:
try:
row = json.loads(row)
except:
continue
if row.get('type', None) != 'ea':
continue
amp_loc = Interval.from_string(row['ac'])
trim_loc = Interval.from_string(row['tc'])
strand = row.get('st')
strand = int(strand) if strand != 'None' else 0
amplicon = Amplicon(
chr=amp_loc.chrom, start=amp_loc.start, end=amp_loc.end, strand=strand,
trim_start = trim_loc.start, trim_end=trim_loc.end,
external_id = row['id'], stats = stats,
offset_allowed=10,
)
amplicons.append(amplicon)
return amplicons
def test_Interval_intersection():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-10964:-1')
i = l1.intersection(l2)
assert i.start == l2.start
assert i.end == l2.end
assert i.chrom == l2.chrom
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr1:10858-11000:-1')
i = l1.intersection(l2)
assert i.start == l2.start
assert i.end == l1.end
assert i.chrom == l2.chrom
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = Interval.from_string('chr2:10858-10964:-1')
i = l1.intersection(l2)
assert i is None
def test_add_border():
l1 = Interval.from_string('chr1:10000-10967:1')
l2 = l1.add_border(upstream=50, downstream=100)
assert l2.start == 10000 - 50
assert l2.end == 10967 + 100
def test_find_minimal_spanning_set():
targets = [
Interval.from_string('chr1:10000-11000:+'),
Interval.from_string('chr1:12000-13000:+')
]
candidates = [
Interval.from_string('chr1:10000-10500:+'),
Interval.from_string('chr1:10100-10600:+'),
Interval.from_string('chr1:10300-11700:+'),
Interval.from_string('chr1:10500-11000:+'),
Interval.from_string('chr1:10700-11100:+'),
]
reads = MinimalSpanningSet(targets, candidates)
n_reads = len(reads.chosen)
print 'chose', reads.chosen
print 'n_reads', n_reads
assert n_reads == 2
targets = [
Interval.from_string('chr1:12000-13000:+')
]
candidates = [
Interval.from_string('chr1:10000-10500:+'),
Interval.from_string('chr1:10100-10600:+'),
Interval.from_string('chr1:10300-11700:+'),
Interval.from_string('chr1:10500-11000:+'),
Interval.from_string('chr1:10700-11100:+'),
]
# this example is constructed into fooling the greedy algorithm
# to choose one too many
reads = MinimalSpanningSet(targets, candidates)
assert len(reads.chosen) == 0
targets = [
Interval.from_string('chr1:12000-13000:+')
]
candidates = [
Interval.from_string('chr1:12000-12500:+'),
Interval.from_string('chr1:12500-13000:+'),
Interval.from_string('chr1:12100-12900:+'),
]
reads = MinimalSpanningSet(targets, candidates)
assert len(reads.chosen) == 2
