def test_negative(self):
transcript = Transcript()
transcript.chrom, transcript.start, transcript.end, transcript.strand = "Chr5", 26575364, 26578163, "-"
transcript.id, transcript.parent = "cufflinks_star_at.23553.1", "cufflinks_star_at.23553"
transcript.add_exons([(26575364, 26575410), (26575495, 26575620),
(26575711, 26575797), (26575885, 26575944),
(26576035, 26576134), (26576261, 26577069),
(26577163, 26577288), (26577378, 26577449),
(26577856, 26578163)])
transcript.finalize()
fasta_seq = self.fasta.fetch(transcript.chrom, transcript.start - 1,
transcript.end)
tr_neg = transcript.copy()
tchecker = TranscriptChecker(tr_neg, fasta_seq, strand_specific=False)
self.assertEqual(tchecker.strand, "-")
self.assertEqual(str(tchecker.fasta_seq.seq), fasta_seq)
tchecker.check_strand()
self.assertEqual(tchecker.strand, "-")
tr_neg = transcript.copy()
tr_neg.strand = "+"
for ss in (False, True):
with self.subTest(ss=ss):
tchecker = TranscriptChecker(tr_neg.copy(),
fasta_seq,
strand_specific=ss)
tchecker.check_strand()
self.assertEqual(tchecker.strand, "-")
def test_check_strand_not_reversed(self):
self.model.strand = "-"
tcheck = TranscriptChecker(self.model, self.model_fasta, strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "-")
self.assertTrue(tcheck.attributes["canonical_on_reverse_strand"])
self.assertTrue(tcheck.suspicious_splicing)
def test_check_reverse_strand(self):
self.model.strand = "-"
tcheck = TranscriptChecker(self.model,
self.model_fasta,
strand_specific=False)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "+")
def test_check_strand_not_reversed(self):
self.model.strand = "-"
tcheck = TranscriptChecker(self.model,
self.model_fasta,
strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "-")
self.assertTrue(tcheck.attributes["canonical_on_reverse_strand"])
self.assertTrue(tcheck.suspicious_splicing)
def test_reference_not_flipped(self):
model = Transcript()
model.chrom, model.start, model.end = "Chr5", 9930, 13235
model.id, model.parent, model.strand = "AT5G01030.1", "AT5G01030", "-"
model.add_exons([(9930, 10172), (10620, 12665), (12797, 13235)])
model.add_exons([(10638, 12665), (12797, 13003)], features="CDS")
model.finalize()
model_fasta = self.fasta["Chr5"][model.start - 1:model.end]
check_model = TranscriptChecker(model, model_fasta, is_reference=True)
check_model.check_strand()
self.assertEqual(check_model.strand, "-")
self.assertGreater(check_model.combined_cds_length, 0)
def test_check_reverse_strand(self):
self.model.strand = "-"
tcheck = TranscriptChecker(self.model,
self.model_fasta,
strand_specific=False)
logger = create_default_logger("test_check_reverse_strand", "DEBUG")
tcheck.logger = logger
self.assertFalse(tcheck.strand_specific)
self.assertFalse(tcheck.is_reference)
self.assertFalse(tcheck.lenient)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "+")
def test_monoexonic(self):
exon = self.gff_lines[1]
transcript_line = self.gff_lines[0]
transcript_line.end = exon.end
model = Transcript(transcript_line)
model.add_exon(exon)
model.finalize()
fasta = self.fasta[model.chrom][model.start - 1:model.end]
tcheck = TranscriptChecker(model.copy(), fasta, strand_specific=False)
tcheck.check_strand()
self.assertIsNone(tcheck.strand)
tcheck = TranscriptChecker(model.copy(), fasta, strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "+")
neg = model.copy()
neg.strand = "-"
tcheck = TranscriptChecker(neg.copy(), fasta, strand_specific=False)
tcheck.check_strand()
self.assertIsNone(tcheck.strand)
tcheck = TranscriptChecker(neg.copy(), fasta, strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "-")
def test_monoexonic(self):
self.model.unfinalize()
for exon in sorted(self.model.exons)[1:]:
self.model.remove_exon(exon)
self.model.finalize()
fasta = self.fasta[self.model.chrom][self.model.start -
1:self.model.end]
tcheck = TranscriptChecker(self.model.copy(),
fasta,
strand_specific=False)
tcheck.check_strand()
self.assertIsNone(tcheck.strand)
tcheck = TranscriptChecker(self.model.copy(),
fasta,
strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "+")
neg = self.model.copy()
neg.strand = "-"
tcheck = TranscriptChecker(neg.copy(), fasta, strand_specific=False)
tcheck.check_strand()
self.assertIsNone(tcheck.strand)
tcheck = TranscriptChecker(neg.copy(), fasta, strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "-")
def test_monoexonic(self):
exon = self.gff_lines[1]
transcript_line = self.gff_lines[0]
transcript_line.end = exon.end
model = Transcript(transcript_line)
model.add_exon(exon)
model.finalize()
fasta = self.fasta[model.chrom][model.start - 1: model.end]
tcheck = TranscriptChecker(model.copy(), fasta, strand_specific=False)
tcheck.check_strand()
self.assertIsNone(tcheck.strand)
tcheck = TranscriptChecker(model.copy(), fasta, strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "+")
neg = model.copy()
neg.strand = "-"
tcheck = TranscriptChecker(neg.copy(), fasta, strand_specific=False)
tcheck.check_strand()
self.assertIsNone(tcheck.strand)
tcheck = TranscriptChecker(neg.copy(), fasta, strand_specific=True)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "-")
def test_negative(self):
gtf_lines = """Chr5 Cufflinks transcript 26575364 26578163 1000 - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";exon_number "1";FPKM "2.9700103727";conf_hi "3.260618";frac "0.732092";cov "81.895309";conf_lo "2.679403";
Chr5 Cufflinks exon 26575364 26575410 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26575495 26575620 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26575711 26575797 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26575885 26575944 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26576035 26576134 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26576261 26577069 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26577163 26577288 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26577378 26577449 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26577856 26578163 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";"""
gtf_lines = [GtfLine(line) for line in gtf_lines.split("\n")]
self.assertEqual(len([_ for _ in gtf_lines if _.header]), 0)
transcript = Transcript(gtf_lines[0])
transcript.add_exons(gtf_lines[1:])
transcript.finalize()
fasta_seq = self.fasta[transcript.chrom][transcript.start -
1:transcript.end]
tr_neg = transcript.copy()
tchecker = TranscriptChecker(tr_neg, fasta_seq, strand_specific=False)
self.assertEqual(tchecker.strand, "-")
self.assertEqual(tchecker.fasta_seq, fasta_seq)
tchecker.check_strand()
self.assertEqual(tchecker.strand, "-")
tr_neg = transcript.copy()
tr_neg.strand = "+"
for ss in (False, True):
with self.subTest(ss=ss):
tchecker = TranscriptChecker(tr_neg.copy(),
fasta_seq,
strand_specific=ss)
tchecker.check_strand()
if ss:
self.assertEqual(tchecker.strand, "+")
self.assertTrue(tchecker.suspicious_splicing)
else:
self.assertEqual(tchecker.strand, "-")
def test_monoexonic_cds(self):
# Chr5 tair10 exon 26584797 26584879 . + . ID=c58_g1_i3.mrna1.19.exon1;Parent=c58_g1_i3.mrna1.19
for strand in ("+", "-"):
with self.subTest(strand=strand):
model = self.model.copy()
model.unfinalize()
[model.remove_exon(exon) for exon in sorted(model.exons[1:])]
exon = model.exons[0]
model.add_exon((exon[0] + 2, exon[1]), feature="CDS")
model.strand = strand
model.finalize()
self.assertTrue(model.is_coding)
fasta = self.fasta[model.chrom][model.start - 1:model.end]
tcheck = TranscriptChecker(model.copy(),
fasta,
strip_faulty_cds=False,
strand_specific=False)
tcheck.check_strand()
self.assertTrue(tcheck.is_coding)
self.assertEqual(tcheck.strand, strand)
def test_negative(self):
gtf_lines = """Chr5 Cufflinks transcript 26575364 26578163 1000 - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";exon_number "1";FPKM "2.9700103727";conf_hi "3.260618";frac "0.732092";cov "81.895309";conf_lo "2.679403";
Chr5 Cufflinks exon 26575364 26575410 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26575495 26575620 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26575711 26575797 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26575885 26575944 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26576035 26576134 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26576261 26577069 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26577163 26577288 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26577378 26577449 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";
Chr5 Cufflinks exon 26577856 26578163 . - . gene_id "cufflinks_star_at.23553";transcript_id "cufflinks_star_at.23553.1";"""
gtf_lines = [GtfLine(line) for line in gtf_lines.split("\n")]
self.assertEqual(len([_ for _ in gtf_lines if _.header]), 0)
transcript = Transcript(gtf_lines[0])
transcript.add_exons(gtf_lines[1:])
transcript.finalize()
fasta_seq = self.fasta[transcript.chrom][transcript.start - 1:transcript.end]
tr_neg = transcript.copy()
tchecker = TranscriptChecker(tr_neg, fasta_seq, strand_specific=False)
self.assertEqual(tchecker.strand, "-")
self.assertEqual(tchecker.fasta_seq, fasta_seq)
tchecker.check_strand()
self.assertEqual(tchecker.strand, "-")
tr_neg = transcript.copy()
tr_neg.strand = "+"
for ss in (False, True):
with self.subTest(ss=ss):
tchecker = TranscriptChecker(tr_neg.copy(), fasta_seq, strand_specific=ss)
tchecker.check_strand()
if ss:
self.assertEqual(tchecker.strand, "+")
self.assertTrue(tchecker.suspicious_splicing)
else:
self.assertEqual(tchecker.strand, "-")
def test_reverse_with_cds_positive(self):
model = Transcript()
model.chrom, model.start, model.end = "Chr5", 9930, 13235
model.id, model.parent, model.strand = "AT5G01030.1", "AT5G01030", "+"
model.add_exons([(9930, 10172), (10620, 12665), (12797, 13235)])
model.add_exons([(10638, 12665), (12797, 13003)], features="CDS")
model.finalize()
model_fasta = self.fasta["Chr5"][model.start - 1:model.end]
check_model = TranscriptChecker(model,
model_fasta,
strip_faulty_cds=True)
check_model.check_strand()
self.assertEqual(check_model.strand, "+")
self.assertGreater(check_model.combined_cds_length, 0)
model.unfinalize()
model.strand = "-"
check_model = TranscriptChecker(model,
model_fasta,
strip_faulty_cds=True)
check_model.check_strand()
self.assertEqual(check_model.strand, "+")
self.assertFalse(check_model.is_coding)
check_model = TranscriptChecker(model,
model_fasta,
strip_faulty_cds=False)
with self.assertRaises(InvalidTranscript):
check_model.check_strand()
def test_reverse_with_cds_negative(self):
model = Transcript()
model.chrom, model.start, model.end = "Chr5", 1251, 5043
model.id, model.parent, model.strand = "AT5G01010.1", "AT5G01010", "-"
model.add_exons([(4765, 5043), (4552, 4679), (4335, 4467),
(4102, 4258), (3927, 4005), (3762, 3802),
(3543, 3659), (3303, 3383), (2872, 2934),
(2748, 2799), (2435, 2509), (1914, 1961),
(1745, 1780), (1572, 1646), (1251, 1459)])
model.add_exons([(4765, 4924), (4552, 4679), (4335, 4467),
(4102, 4258), (3927, 4005), (3762, 3802),
(3543, 3659), (3303, 3383), (2872, 2934),
(2748, 2799), (2435, 2509), (1914, 1961),
(1745, 1780), (1572, 1646), (1388, 1459)],
features="CDS")
model.finalize()
model_fasta = self.fasta["Chr5"][model.start - 1:model.end]
check_model = TranscriptChecker(model,
model_fasta,
strip_faulty_cds=True)
check_model.check_strand()
self.assertEqual(check_model.strand, "-")
self.assertGreater(check_model.combined_cds_length, 0)
model.unfinalize()
model.strand = "+"
check_model = TranscriptChecker(model,
model_fasta,
strip_faulty_cds=True)
check_model.check_strand()
self.assertEqual(check_model.strand, "-")
self.assertFalse(check_model.is_coding)
check_model = TranscriptChecker(model,
model_fasta,
strip_faulty_cds=False)
# Check that if we want to keep the CDS, this will raise an error
with self.assertRaises(InvalidTranscript):
check_model.check_strand()
def create_transcript(lines,
fasta_seq,
start,
end,
lenient=False,
strand_specific=False,
canonical_splices=(("GT", "AG"), ("GC", "AG"), ("AT",
"AC")),
logger=None):
"""Function to create the checker.
:param lines: all the exon lines for an object
:type lines: dict
:param fasta_seq: genomic sequence of the transcript
:param start: start position for the transcript
:type start: int
:param end: end position for the transcript
:type end: int
:type lenient: bool
:type strand_specific: bool
:param canonical_splices: the splices considered as canonical for the species.
:type canonical_splices: list[tuple]
:param logger: optional logger to use during processing.
:rtype: (None|TranscriptChecker)
"""
if logger is None:
logger = create_null_logger("checker")
logger.debug("Starting with %s", lines["tid"])
try:
transcript_line = Transcript()
transcript_line.chrom = lines["chrom"]
if "source" in lines:
transcript_line.source = lines["source"]
transcript_line.strand = lines["strand"]
transcript_line.attributes.update(lines["attributes"])
transcript_line.feature = "transcript"
transcript_line.start, transcript_line.end = sorted([start, end])
transcript_line.logger = logger
assert lines["tid"] is not None, lines
transcript_line.id = lines["tid"]
transcript_line.parent = lines["parent"]
for feature in lines["features"]:
transcript_line.add_exons(lines["features"][feature],
features=feature)
transcript_object = TranscriptChecker(
transcript_line,
fasta_seq,
lenient=lenient,
strand_specific=strand_specific,
canonical_splices=canonical_splices,
logger=logger)
logger.debug("Finished adding exon lines to %s", lines["tid"])
transcript_object.finalize()
transcript_object.check_strand()
except exceptions.IncorrectStrandError:
logger.info(
"Discarded %s because of incorrect fusions of splice junctions",
lines["tid"])
# logger.exception(exc)
transcript_object = None
except exceptions.InvalidTranscript:
logger.info("Discarded generically invalid transcript %s",
lines["tid"])
transcript_object = None
except KeyboardInterrupt:
raise KeyboardInterrupt
except Exception as exc:
logger.exception(exc)
transcript_object = None
logger.debug("Finished with %s", lines["tid"])
return transcript_object
def create_transcript(lines,
fasta_seq,
start,
end,
lenient=False,
strand_specific=False,
canonical_splices=(("GT", "AG"),
("GC", "AG"),
("AT", "AC")),
logger=None):
"""Function to create the checker.
:param lines: all the exon lines for an object
:type lines: dict
:param fasta_seq: genomic sequence of the transcript
:param start: start position for the transcript
:type start: int
:param end: end position for the transcript
:type end: int
:type lenient: bool
:type strand_specific: bool
:param canonical_splices: the splices considered as canonical for the species.
:type canonical_splices: list[tuple]
:param logger: optional logger to use during processing.
:rtype: (None|TranscriptChecker)
"""
if logger is None:
logger = create_null_logger("checker")
logger.debug("Starting with %s", lines["tid"])
try:
transcript_line = Transcript()
transcript_line.chrom = lines["chrom"]
if "source" in lines:
transcript_line.source = lines["source"]
transcript_line.strand = lines["strand"]
transcript_line.attributes.update(lines["attributes"])
transcript_line.feature = "transcript"
transcript_line.start, transcript_line.end = sorted([start, end])
transcript_line.logger = logger
assert lines["tid"] is not None, lines
transcript_line.id = lines["tid"]
transcript_line.parent = lines["parent"]
for feature in lines["features"]:
transcript_line.add_exons(lines["features"][feature],
features=feature)
transcript_object = TranscriptChecker(transcript_line,
fasta_seq,
lenient=lenient,
strand_specific=strand_specific,
canonical_splices=canonical_splices,
logger=logger)
logger.debug("Finished adding exon lines to %s", lines["tid"])
transcript_object.finalize()
transcript_object.check_strand()
except exceptions.IncorrectStrandError:
logger.info("Discarded %s because of incorrect fusions of splice junctions",
lines["tid"])
# logger.exception(exc)
transcript_object = None
except exceptions.InvalidTranscript:
logger.info("Discarded generically invalid transcript %s",
lines["tid"])
transcript_object = None
except KeyboardInterrupt:
raise KeyboardInterrupt
except Exception as exc:
logger.exception(exc)
transcript_object = None
logger.debug("Finished with %s", lines["tid"])
return transcript_object
def create_transcript(lines,
fasta_seq,
start,
end,
lenient=False,
is_reference=False,
strand_specific=False,
canonical_splices=(("GT", "AG"), ("GC", "AG"), ("AT",
"AC")),
force_keep_cds=False,
logger=None):
"""Function to create the checker.
:param lines: all the exon lines for an object
:type lines: dict
:param fasta_seq: genomic sequence of the transcript
:param start: start position for the transcript
:type start: int
:param end: end position for the transcript
:type end: int
:type lenient: bool
:type strand_specific: bool
:param canonical_splices: the splices considered as canonical for the species.
:type canonical_splices: list[tuple]
:param force_keep_cds: boolean. If set to true, coding transcripts that would be flipped are instead excluded.
The intention is that this flag will mirror strip_cds.
:type force_keep_cds: bool
:param logger: optional logger to use during processing.
:param is_reference: boolean. If set, the transcript's strand will not be checked.
:rtype: (None|TranscriptChecker)
"""
if logger is None:
logger = create_null_logger()
if "tid" not in lines:
logger.error("Lines datastore lacks the transcript ID. Exiting.")
return None
try:
logger.debug("Starting with %s", lines["tid"])
transcript_line = Transcript()
transcript_line.chrom = lines["chrom"]
if "source" in lines:
transcript_line.source = lines["source"]
transcript_line.strand = lines["strand"]
transcript_line.attributes.update(lines["attributes"])
transcript_line.feature = "transcript"
transcript_line.start, transcript_line.end = sorted([start, end])
transcript_line.logger = logger
assert lines["tid"] is not None, lines
transcript_line.id = lines["tid"]
transcript_line.parent = lines["parent"]
for feature in lines["features"]:
coords, phases = [], []
for feat in lines["features"][feature]:
try:
assert isinstance(
feat, (list, tuple)) and 2 <= len(feat) <= 3, feat
except AssertionError:
raise exceptions.InvalidTranscript("Invalid feature")
coords.append((feat[0], feat[1]))
if len(feat) == 3 and feat[2] in (0, 1, 2, None):
phases.append(feat[2])
else:
phases.append(None)
try:
assert len(phases) == len(coords)
except AssertionError:
raise exceptions.InvalidTranscript("Invalid phases/coords")
transcript_line.add_exons(coords, features=feature, phases=phases)
transcript_object = TranscriptChecker(
transcript_line,
fasta_seq,
lenient=lenient,
strand_specific=strand_specific,
canonical_splices=canonical_splices,
force_keep_cds=force_keep_cds,
is_reference=is_reference,
logger=logger)
logger.debug("Finished adding exon lines to %s", lines["tid"])
transcript_object.finalize()
transcript_object.check_strand()
transcript_object.check_orf()
except exceptions.IncorrectStrandError:
logger.info(
"Discarded %s because of incorrect fusions of splice junctions",
lines["tid"])
# logger.exception(exc)
transcript_object = None
except exceptions.InvalidTranscript as exc:
logger.info(
"Discarded generically invalid transcript %s, exception: %s",
lines["tid"], exc)
transcript_object = None
except AssertionError as exc:
logger.info("Validation failed on %s, assertion failure: %s",
lines["tid"], exc)
transcript_object = None
except KeyboardInterrupt:
raise KeyboardInterrupt
except Exception as exc:
logger.exception(exc)
transcript_object = None
return transcript_object
def test_check_reverse_strand(self):
self.model.strand = "-"
tcheck = TranscriptChecker(self.model, self.model_fasta, strand_specific=False)
tcheck.check_strand()
self.assertEqual(tcheck.strand, "+")
