def test_introns():
t = Transfrag(chrom='chrTest', strand=Strand.POS,
exons=[Exon(0, 10), Exon(20, 30), Exon(40, 50)])
introns = list(t.iterintrons())
assert len(introns) == 2
assert introns[0] == (10, 20)
assert introns[1] == (30, 40)
def make_ramp(strand, sign=1):
transfrags = []
chrom = 'chr1'
start = 1000
end = 1220
change_expr = 0.0
base_expr = 0.0
# "flat" part of expression landscape
expr = 1.0
for i in xrange(0, 50):
t = Transfrag(chrom=chrom, strand=strand,
_id='T1.%d' % i, sample_id='S%d' % i,
expr=expr, is_ref=False,
exons=[Exon(start, end)])
transfrags.append(t)
change_expr += expr
base_expr += expr
# "changing" area
i = 0
expr = 10.0
for pos in range(1100, 1120):
left, right = (start, pos) if sign < 0 else (pos, end)
t = Transfrag(chrom=chrom, strand=strand,
_id='T2.%d' % i, sample_id='S%d' % i,
expr=expr, is_ref=False,
exons=[Exon(left, right)])
transfrags.append(t)
change_expr += expr
i += 1
return chrom, start, end, strand, change_expr, base_expr, transfrags
def test_splices():
t = Transfrag(chrom='chrTest',
strand=Strand.POS,
exons=[Exon(0, 10), Exon(20, 30),
Exon(40, 50)])
splice_sites = frozenset(t.itersplices())
assert len(splice_sites) == 4
assert splice_sites == frozenset((10, 20, 30, 40))
def test_introns():
t = Transfrag(chrom='chrTest',
strand=Strand.POS,
exons=[Exon(0, 10), Exon(20, 30),
Exon(40, 50)])
introns = list(t.iterintrons())
assert len(introns) == 2
assert introns[0] == (10, 20)
assert introns[1] == (30, 40)
def read_single_locus(filename, guided_strand=False):
loci = read_gtf(filename)
assert len(loci) == 1
interval, gtf_lines = loci[0]
t_dict = Transfrag.parse_gtf(gtf_lines)
locus = Locus.create(t_dict.values(), guided_strand=guided_strand)
return t_dict, locus
def parse_gtf_locus(locus, gtf_fileh):
genome_id_str = '%s:%d-%d' % (locus.chrom, locus.start, locus.end)
logging.debug('%s locus: %s features: %d' %
(genome_id_str, locus.name, locus.num_lines))
# fast-forward to 'filepos'
gtf_fileh.seek(locus.filepos)
# read 'num_lines' lines from file and parse into transfrag objects
t_dict = Transfrag.parse_gtf(next(gtf_fileh)
for x in xrange(locus.num_lines))
return t_dict.values()
def parse_gtf_locus(locus, gtf_fileh):
genome_id_str = '%s:%d-%d' % (locus.chrom, locus.start, locus.end)
logging.debug('%s locus: %s features: %d' %
(genome_id_str, locus.name, locus.num_lines))
# fast-forward to 'filepos'
gtf_fileh.seek(locus.filepos)
# read 'num_lines' lines from file and parse into transfrag objects
t_dict = Transfrag.parse_gtf(
next(gtf_fileh) for x in xrange(locus.num_lines))
return t_dict.values()
def parse_locus(locus, fh):
genome_id_str = '%s:%d-%d' % (locus.chrom, locus.start, locus.end)
logging.debug('LocusIndex: %s coords: %s transfrags: %d' %
(locus.name, genome_id_str, locus.num_lines))
# fast-forward to 'filepos'
fh.seek(locus.filepos)
# parse 'num_lines' from file into Transfrag objects
transfrags = []
for i in xrange(locus.num_lines):
transfrags.append(Transfrag.from_bed(fh.next()))
return transfrags
def parse_locus(locus, fh):
genome_id_str = '%s:%d-%d' % (locus.chrom, locus.start, locus.end)
logging.debug('LocusIndex: %s coords: %s transfrags: %d' %
(locus.name, genome_id_str, locus.num_lines))
# fast-forward to 'filepos'
fh.seek(locus.filepos)
# parse 'num_lines' from file into Transfrag objects
transfrags = []
for i in xrange(locus.num_lines):
transfrags.append(Transfrag.from_bed(fh.next()))
return transfrags
def main():
logging.basicConfig(level=logging.DEBUG)
parser = argparse.ArgumentParser()
parser.add_argument('genome_fasta_file')
parser.add_argument('bed_file')
args = parser.parse_args()
# check args
if not os.path.exists(args.genome_fasta_file):
parser.error('genome fasta file %s not found' % args.genome_fasta_file)
if not os.path.exists(args.bed_file):
parser.error('bed file %s not found' % args.bed_file)
logging.info('genome fasta file: %s' % args.genome_fasta_file)
logging.info('bed file: %s' % args.bed_file)
# process bed file to get junctions
logging.info('Reading Junctions')
splice_juncs = set()
fasta_fh = FastaFile(args.genome_fasta_file)
with open(args.bed_file) as bed_fh:
for line in bed_fh:
t = Transfrag.from_bed(line)
if t.chrom not in fasta_fh:
continue
for start, end in t.iterintrons():
splice_juncs.add((t.chrom, start, end, t.strand))
logging.info('Read %d Junctions' % (len(splice_juncs)))
logging.info('Profiling Splice Motifs')
motif_counter = Counter()
for chrom, start, end, strand in splice_juncs:
s = fasta_fh.fetch(chrom, start, start + 2)
s += fasta_fh.fetch(chrom, end - 2, end)
if strand == Strand.NEG:
s = dna_reverse_complement(s)
motif_counter[s] += 1
fasta_fh.close()
# report statistics
total = sum(motif_counter.values())
print '\t'.join(['motif', 'count', 'frac'])
for motif, count in motif_counter.most_common():
print '\t'.join([motif, str(count), str(float(count) / total)])
logging.info('Done')
def test_split_transfrag():
loci = read_gtf("splice_sites.gtf")
interval, gtf_lines = loci[0]
t_dict = Transfrag.parse_gtf(gtf_lines)
sg = SpliceGraph.create(t_dict.values())
boundaries = array("i", sg._find_node_boundaries())
# check nodes
t = t_dict["A"]
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((10, 100), (200, 250), (250, 300), (400, 525))
t = t_dict["B"]
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((10, 100), (250, 300), (400, 525))
t = t_dict["C"]
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((150, 200), (200, 250), (250, 300), (400, 525))
t = t_dict["D"]
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((375, 400), (400, 525))
def test_split_transfrag():
loci = read_gtf('splice_sites.gtf')
interval, gtf_lines = loci[0]
t_dict = Transfrag.parse_gtf(gtf_lines)
sg = SpliceGraph.create(t_dict.values())
boundaries = array('i', sg._find_node_boundaries())
# check nodes
t = t_dict['A']
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((10, 100), (200, 250), (250, 300), (400, 525))
t = t_dict['B']
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((10, 100), (250, 300), (400, 525))
t = t_dict['C']
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((150, 200), (200, 250), (250, 300), (400, 525))
t = t_dict['D']
nodes = tuple(split_transfrag(t, boundaries))
assert nodes == ((375, 400), (400, 525))
def test_create_locus():
loci = read_gtf('splice_sites.gtf')
assert len(loci) == 1
interval, gtf_lines = loci[0]
assert interval == ('chr1', 10, 525)
t_dict = Transfrag.parse_gtf(gtf_lines)
locus = Locus.create(t_dict.values())
assert locus.chrom == 'chr1'
assert locus.start == 10
assert locus.end == 525
a = locus.get_expr_data(49, 51, Strand.POS)
assert np.array_equal(a, [1.0, 2.0])
a = locus.get_expr_data(150, 151, Strand.POS)
assert np.array_equal(a, [1.0])
a = locus.get_expr_data(499, 501, Strand.POS)
assert np.array_equal(a, [3.0, 1.0])
with pytest.raises(TacoError):
locus.get_expr_data(5, 15, Strand.POS)
with pytest.raises(TacoError):
locus.get_expr_data(520, 530, Strand.POS)
def test_multi_strand1():
# read gtf and test basic values
loci = read_gtf("multi_strand1.gtf")
assert len(loci) == 1
interval, gtf_lines = loci[0]
assert interval == ("chr1", 100, 1000)
t_dict = Transfrag.parse_gtf(gtf_lines)
assert len(t_dict) == 5
locus = Locus.create(t_dict.values())
assert locus.chrom == "chr1"
assert locus.start == 100
assert locus.end == 1000
# raise exception when creating with multiple strands
with pytest.raises(TacoError):
SpliceGraph.create(t_dict.values())
transfrags_pos = locus.get_transfrags(Strand.POS)
transfrags_neg = locus.get_transfrags(Strand.NEG)
sgpos = SpliceGraph.create(transfrags_pos)
sgneg = SpliceGraph.create(transfrags_neg)
# test
assert sgpos.chrom == "chr1"
assert sgpos.start == 100
assert sgpos.end == 650
assert sgpos.strand == Strand.POS
assert sgpos.ref_start_sites == [150]
assert sgpos.ref_stop_sites == [600]
with pytest.raises(TacoError):
sgpos.get_expr_data(90, 110)
with pytest.raises(TacoError):
sgpos.get_expr_data(650, 655)
assert np.array_equal(sgpos.get_expr_data(100, 105), np.ones(5))
assert sgneg.chrom == "chr1"
assert sgneg.start == 350
assert sgneg.end == 1000
assert sgneg.strand == Strand.NEG
assert sgneg.ref_start_sites == [1000]
assert sgneg.ref_stop_sites == [350]
with pytest.raises(TacoError):
sgneg.get_expr_data(340, 350)
with pytest.raises(TacoError):
sgneg.get_expr_data(1000, 1010)
assert np.array_equal(sgneg.get_expr_data(400, 405), np.ones(5))
assert np.array_equal(sgneg.get_expr_data(945, 950), np.zeros(5))
assert np.array_equal(sgneg.get_expr_data(950, 955), np.ones(5))
assert np.array_equal(sgneg.get_expr_data(980, 985), np.zeros(5))
# test locus boundaries
bpos = tuple(sgpos._find_node_boundaries())
assert bpos == tuple((100, 200, 300, 400, 650))
bneg = tuple(sgneg._find_node_boundaries())
assert bneg == tuple((350, 400, 500, 950, 980, 1000))
# added guided ends/assembly to use boundaries from reference
lpos = SpliceGraph.create(transfrags_pos, guided_ends=True, guided_assembly=True)
bpos = tuple(lpos._find_node_boundaries())
assert bpos == tuple((100, 150, 200, 300, 400, 500, 600, 650))
lneg = SpliceGraph.create(transfrags_neg, guided_ends=True, guided_assembly=True)
bneg = tuple(lneg._find_node_boundaries())
assert bneg == tuple((350, 400, 500, 750, 900, 950, 980, 1000))
def test_splices():
t = Transfrag(chrom='chrTest', strand=Strand.POS,
exons=[Exon(0, 10), Exon(20, 30), Exon(40, 50)])
splice_sites = frozenset(t.itersplices())
assert len(splice_sites) == 4
assert splice_sites == frozenset((10, 20, 30, 40))
def test_multi_strand1():
# read gtf and test basic values
loci = read_gtf('multi_strand1.gtf')
assert len(loci) == 1
interval, gtf_lines = loci[0]
assert interval == ('chr1', 100, 1000)
t_dict = Transfrag.parse_gtf(gtf_lines)
assert len(t_dict) == 5
locus = Locus.create(t_dict.values())
assert locus.chrom == 'chr1'
assert locus.start == 100
assert locus.end == 1000
# raise exception when creating with multiple strands
with pytest.raises(TacoError):
SpliceGraph.create(t_dict.values())
transfrags_pos = locus.get_transfrags(Strand.POS)
transfrags_neg = locus.get_transfrags(Strand.NEG)
sgpos = SpliceGraph.create(transfrags_pos)
sgneg = SpliceGraph.create(transfrags_neg)
# test
assert sgpos.chrom == 'chr1'
assert sgpos.start == 100
assert sgpos.end == 650
assert sgpos.strand == Strand.POS
assert sgpos.ref_start_sites == [150]
assert sgpos.ref_stop_sites == [600]
with pytest.raises(TacoError):
sgpos.get_expr_data(90, 110)
with pytest.raises(TacoError):
sgpos.get_expr_data(650, 655)
assert np.array_equal(sgpos.get_expr_data(100, 105), np.ones(5))
assert sgneg.chrom == 'chr1'
assert sgneg.start == 350
assert sgneg.end == 1000
assert sgneg.strand == Strand.NEG
assert sgneg.ref_start_sites == [1000]
assert sgneg.ref_stop_sites == [350]
with pytest.raises(TacoError):
sgneg.get_expr_data(340, 350)
with pytest.raises(TacoError):
sgneg.get_expr_data(1000, 1010)
assert np.array_equal(sgneg.get_expr_data(400, 405), np.ones(5))
assert np.array_equal(sgneg.get_expr_data(945, 950), np.zeros(5))
assert np.array_equal(sgneg.get_expr_data(950, 955), np.ones(5))
assert np.array_equal(sgneg.get_expr_data(980, 985), np.zeros(5))
# test locus boundaries
bpos = tuple(sgpos._find_node_boundaries())
assert bpos == tuple((100, 200, 300, 400, 650))
bneg = tuple(sgneg._find_node_boundaries())
assert bneg == tuple((350, 400, 500, 950, 980, 1000))
# added guided ends/assembly to use boundaries from reference
lpos = SpliceGraph.create(transfrags_pos,
guided_ends=True,
guided_assembly=True)
bpos = tuple(lpos._find_node_boundaries())
assert bpos == tuple((100, 150, 200, 300, 400, 500, 600, 650))
lneg = SpliceGraph.create(transfrags_neg,
guided_ends=True,
guided_assembly=True)
bneg = tuple(lneg._find_node_boundaries())
assert bneg == tuple((350, 400, 500, 750, 900, 950, 980, 1000))