def test_intron_annotation(self):
"""sequences annotated with Introns should return correct seq"""
for symbol, stable_id, rank, exp_seq5, exp_seq3 in [
('IL2', 'ENST00000226730', 1, 'gtaagtatat', 'actttcttag'),
('IL13', 'ENST00000304506', 3, 'gtaaggcatc', 'tgtcctgcag')]:
gene = asserted_one(self.human.getGenesMatching(Symbol=symbol))
seq = gene.getAnnotatedSeq(feature_types='gene')
intron = asserted_one(seq.getAnnotationsMatching('intron',
'%s-%d'%(stable_id, rank)))
intron_seq = str(seq.getRegionCoveringAll(intron).getSlice())
self.assertEqual(intron_seq[:10], exp_seq5.upper())
self.assertEqual(intron_seq[-10:], exp_seq3.upper())
def test_intron_number(self):
"""number of introns should be correct"""
for gene_id, transcript_id, exp_number in [
('ENSG00000227268', 'ENST00000445946', 0),
('ENSG00000132199', 'ENST00000319815', 8),
('ENSG00000132199', 'ENST00000383578', 15)]:
gene = asserted_one(self.human.getGenesMatching(StableId=gene_id))
transcript = asserted_one(
[t for t in gene.Transcripts if t.StableId==transcript_id])
if exp_number == 0:
self.assertEqual(transcript.Introns, None)
else:
self.assertEqual(len(transcript.Introns), exp_number)
def test_intron_annotation(self):
"""sequences annotated with Introns should return correct seq"""
for symbol, stable_id, rank, exp_seq5, exp_seq3 in [
('IL2', 'ENST00000226730', 1, 'gtaagtatat', 'actttcttag'),
('IL13', 'ENST00000304506', 3, 'gtaaggcatc', 'tgtcctgcag')
]:
gene = asserted_one(self.human.getGenesMatching(Symbol=symbol))
seq = gene.getAnnotatedSeq(feature_types='gene')
intron = asserted_one(
seq.getAnnotationsMatching('intron',
'%s-%d' % (stable_id, rank)))
intron_seq = str(seq.getRegionCoveringAll(intron).getSlice())
self.assertEqual(intron_seq[:10], exp_seq5.upper())
self.assertEqual(intron_seq[-10:], exp_seq3.upper())
def test_intron_number(self):
"""number of introns should be correct"""
for gene_id, transcript_id, exp_number in [
('ENSG00000227268', 'ENST00000445946', 0),
('ENSG00000132199', 'ENST00000583771', 5),
('ENSG00000132199', 'ENST00000340116', 14)
]:
gene = asserted_one(self.human.getGenesMatching(StableId=gene_id))
transcript = asserted_one(
[t for t in gene.Transcripts if t.StableId == transcript_id])
if exp_number == 0:
self.assertEqual(transcript.Introns, None)
else:
self.assertEqual(len(transcript.Introns), exp_number)
def _get_exon_record(self):
# this will be called by _Region parent class to make the location
exon_table = self.db.getTable('exon')
query = sql.select([exon_table], exon_table.c.exon_id == self.exon_id)
records = query.execute()
record = asserted_one(records.fetchall())
self._table_rows['exon'] = record
def _get_exon_stable_id_record(self):
exon_stable_id_table = self.db.getTable('exon_stable_id')
query = sql.select([exon_stable_id_table.c.stable_id],
exon_stable_id_table.c.exon_id == self.exon_id)
records = query.execute()
record = asserted_one(records.fetchall())
self._table_rows['exon_stable_id'] = record
def _get_flanking_seq_data(self):
# maps to flanking_sequence through variation_feature_id
# if this fails, we grab from genomic sequence
variation_id = self._table_rows['variation_feature']['variation_id']
flanking_seq_table = self.flanking_sequence_table
query = sql.select([flanking_seq_table],
flanking_seq_table.c.variation_id == variation_id)
record = asserted_one(query.execute())
self._table_rows['flanking_sequence'] = record
up_seq = record['up_seq']
down_seq = record['down_seq']
# the following two lines are because -- wait for it -- someone has
# entered the string 'NULL' instead of NULL in the MySQL tables!!!
up_seq = [up_seq, None][up_seq == 'NULL']
down_seq = [down_seq, None][down_seq == 'NULL']
seqs = dict(up=up_seq, down=down_seq)
for name, seq in seqs.items():
if seq is not None:
seq = DNA.makeSequence(seq)
else:
resized = [(-301, -1), (1, 301)][name == 'down']
if self.Location.Strand == -1:
resized = [(1, 301), (-301, -1)][name == 'down']
flank = self.Location.resized(*resized)
flanking = self.genome.getRegion(region=flank)
seq = flanking.Seq
seqs[name] = seq
self._cached[('FlankingSeq')] = (seqs['up'][-300:],seqs['down'][:300])
def _get_simple_features(self, db, klass, target_coord, query_coord,
where_feature):
"""returns feature_type records for the query_coord from the
simple_feature table. The returned coord is referenced to
target_coord. At present, only CpG islands being queried."""
simple_feature_table = db.getTable('simple_feature')
feature_types = ['CpGisland']
feature_type_ids=[str(self._feature_type_ids.get(f)) for f in feature_types]
# fix the following
query = sql.select([simple_feature_table],
sql.and_(simple_feature_table.c.analysis_id.in_(feature_type_ids),
simple_feature_table.c.seq_region_id == query_coord.seq_region_id))
query = location_query(simple_feature_table,query_coord.EnsemblStart,
query_coord.EnsemblEnd, query=query,
where=where_feature)
records = query.execute()
for record in records:
coord = Coordinate(self, CoordName=query_coord.CoordName,
Start=record['seq_region_start'],
End = record['seq_region_end'],
seq_region_id=record['seq_region_id'],
Strand = record['seq_region_strand'],
ensembl_coord=True)
if query_coord.CoordName != target_coord.CoordName:
coord = asserted_one(get_coord_conversion(coord, target_coord.CoordType, self.CoreDb))[1]
# coord = coord.makeRelativeTo(query_coord) #TODO: fix here if query_coord and target_coord have different coordName
# coord = coord.makeRelativeTo(target_coord, False)
yield klass(self, db, Location=coord, Score=record['score'])
def _get_repeat_features(self, db, klass, target_coord, query_coord, where_feature):
"""returns Repeat region instances"""
# we build repeats using coordinates from repeat_feature table
# the repeat_consensus_id is required to get the repeat name, class
# and type
repeat_feature_table = db.getTable("repeat_feature")
query = sql.select([repeat_feature_table], repeat_feature_table.c.seq_region_id == query_coord.seq_region_id)
query = location_query(
repeat_feature_table, query_coord.EnsemblStart, query_coord.EnsemblEnd, query=query, where=where_feature
)
for record in query.execute():
coord = Coordinate(
self,
CoordName=query_coord.CoordName,
Start=record["seq_region_start"],
End=record["seq_region_end"],
seq_region_id=record["seq_region_id"],
Strand=record["seq_region_strand"],
ensembl_coord=True,
)
if query_coord.CoordName != target_coord.CoordName:
coord = asserted_one(get_coord_conversion(coord, target_coord.CoordType, self.CoreDb))[1]
# coord = coord.makeRelativeTo(query_coord) #TODO: fix here if query_coord and target_coord have different coordName
# coord = coord.makeRelativeTo(target_coord, False)
yield klass(self, db, Location=coord, Score=record["score"], data=record)
def _get_repeat_features(self, db, klass, target_coord, query_coord,
where_feature):
"""returns Repeat region instances"""
# we build repeats using coordinates from repeat_feature table
# the repeat_consensus_id is required to get the repeat name, class
# and type
repeat_feature_table = db.getTable('repeat_feature')
query = sql.select(
[repeat_feature_table],
repeat_feature_table.c.seq_region_id == query_coord.seq_region_id)
query = location_query(repeat_feature_table,
query_coord.EnsemblStart,
query_coord.EnsemblEnd,
query=query,
where=where_feature)
for record in query.execute():
coord = Coordinate(self,
CoordName=query_coord.CoordName,
Start=record['seq_region_start'],
End=record['seq_region_end'],
seq_region_id=record['seq_region_id'],
Strand=record['seq_region_strand'],
ensembl_coord=True)
if query_coord.CoordName != target_coord.CoordName:
coord = asserted_one(
get_coord_conversion(coord, target_coord.CoordType,
self.CoreDb))[1]
# coord = coord.makeRelativeTo(query_coord) #TODO: fix here if query_coord and target_coord have different coordName
# coord = coord.makeRelativeTo(target_coord, False)
yield klass(self,
db,
Location=coord,
Score=record['score'],
data=record)
def _get_gene_features(self, db, klass, target_coord, query_coord,
where_feature):
"""returns all genes"""
xref_table = [None, db.getTable('xref')][db.Type == 'core']
gene_table = db.getTable('gene')
# after release 65, the gene_id_table is removed. The following is to maintain
# support for earlier releases.
if self.GeneralRelease >= 65:
gene_id_table = None
else:
gene_id_table = db.getTable('gene_stable_id')
# note gene records are at chromosome, not contig, level
condition = gene_table.c.seq_region_id == query_coord.seq_region_id
query = self._build_gene_query(db, condition, gene_table, gene_id_table, xref_table)
query = location_query(gene_table, query_coord.EnsemblStart,
query_coord.EnsemblEnd, query=query, where=where_feature)
for record in query.execute():
new = Coordinate(self, CoordName=query_coord.CoordName,
Start=record['seq_region_start'],
End = record['seq_region_end'],
Strand = record['seq_region_strand'],
seq_region_id=record['seq_region_id'],
ensembl_coord=True)
if query_coord.CoordName != target_coord.CoordName:
coord = asserted_one(get_coord_conversion(coord, target_coord.CoordType, self.CoreDb))[1]
# TODO: check coord, used 'new' here. where is coord (above line) used?
gene = klass(self, db, Location=new, data=record)
yield gene
def _get_seq_region_record(self, seq_region_id):
# should this be on a parent class? or a generic function in assembly?
seq_region_table = self.db.getTable('seq_region')
query = sql.select([seq_region_table],
seq_region_table.c.seq_region_id == seq_region_id)
record = asserted_one(query.execute())
return record
def _get_transcript_stable_id_record(self):
table_name = self._attr_ensembl_table_map['StableId']
if table_name in self._table_rows:
return
transcript_id = self.transcript_id
table = self.db.getTable(table_name)
query = sql.select([table], table.c.transcript_id == transcript_id)
record = asserted_one(query.execute())
self._table_rows[table_name] = record
def _get_cpg_island_analysis_id(self):
analysis_description_table = self.genome.CoreDb.getTable("analysis_description")
query = sql.select(
[analysis_description_table.c.analysis_id], analysis_description_table.c.display_label.like("%CpG%")
)
record = asserted_one(query.execute())
self._table_rows["analysis_description"] = record
quoted_limited = lambda x: DisplayString(x, with_quotes=True, num_words=2)
self._populate_cache_from_record([("CpGisland", "analysis_id", quoted_limited)], "analysis_description")
def _get_cigar_record(self):
genomic_align_table = \
self.parent.compara.ComparaDb.getTable('genomic_align')
query = sql.select([genomic_align_table.c.cigar_line],
genomic_align_table.c.genomic_align_id == \
self._cached['genomic_align_id'])
record = asserted_one(query.execute())
self._cached['cigar_line'] = record['cigar_line']
return record
def _get_cigar_record(self):
genomic_align_table = \
self.parent.compara.ComparaDb.getTable('genomic_align')
query = sql.select([genomic_align_table.c.cigar_line],
genomic_align_table.c.genomic_align_id == \
self._cached['genomic_align_id'])
record = asserted_one(query.execute())
self._cached['cigar_line'] = record['cigar_line']
return record
def _get_seq_region_id(self, CoordName):
"""returns the seq_region_id for the provided CoordName"""
seq_region_table = self.CoreDb.getTable('seq_region')
coord_systems = CoordSystem(core_db=self.CoreDb)
coord_system_ids = [k for k in coord_systems if type(k) not in (unicode, str)]
record = sql.select([seq_region_table.c.seq_region_id],
sql.and_(seq_region_table.c.name == CoordName,
seq_region_table.c.coord_system_id.in_(coord_system_ids)))
record = asserted_one(record.execute().fetchall())
return record['seq_region_id']
def _get_seq_region_id(self, CoordName):
"""returns the seq_region_id for the provided CoordName"""
seq_region_table = self.CoreDb.getTable('seq_region')
coord_systems = CoordSystem(core_db=self.CoreDb)
coord_system_ids = [k for k in coord_systems if not isinstance(k, str)]
record = sql.select(
[seq_region_table.c.seq_region_id],
sql.and_(seq_region_table.c.name == CoordName,
seq_region_table.c.coord_system_id.in_(coord_system_ids)))
record = asserted_one(record.execute().fetchall())
return record['seq_region_id']
def _get_translation_record(self):
transcript_id = self.transcript_id
translation_table = self.db.getTable('translation')
query = sql.select([translation_table],
translation_table.c.transcript_id == transcript_id)
try:
record = asserted_one(query.execute())
except NoItemError:
self._set_null_values(['TranslatedExons'], 'translation')
return
self._table_rows['translation'] = record
def test_intron(self):
"""should get correct Intron sequence, regardless of strand"""
# IL2 is on - strand, IL13 is on + strand, both have three introns
IL2_exp_introns = [
(1, 122456203, 122456293, 'gtaagtatat', 'actttcttag'),
(2, 122453853, 122456143, 'gtaagtacaa', 'attattctag'),
(3, 122451862, 122453709, 'gtaaggcatt', 'tcttttatag')
]
IL13_exp_introns = [
(1, 132658360, 132659417, 'gtgagtgtcg', 'gctcccacag'),
(2, 132659471, 132659723, 'gtaaggacct', 'ctccccacag'),
(3, 132659828, 132660174, 'gtaaggcatc', 'tgtcctgcag')
]
for symbol, stable_id, exp_introns in [
('IL2', 'ENST00000226730', IL2_exp_introns),
('IL13', 'ENST00000304506', IL13_exp_introns)
]:
gene = asserted_one(self.human.getGenesMatching(Symbol=symbol))
strand = gene.Location.Strand
transcript = asserted_one(
[t for t in gene.Transcripts if t.StableId == stable_id])
introns = transcript.Introns
self.assertEqual(len(introns), len(exp_introns))
idx = 0
for intron in introns:
loc = intron.Location
start, end = loc.Start, loc.End
seq = str(intron.Seq)
exp_rank, exp_start, exp_end, exp_seq5, \
exp_seq3 = exp_introns[idx]
self.assertEqual(loc.Strand, strand)
# test the order using rank
self.assertEqual(intron.Rank, exp_rank)
# test position
self.assertEqual(start, exp_start)
self.assertEqual(end, exp_end)
# test sequence
self.assertEqual(seq[:10], exp_seq5.upper())
self.assertEqual(seq[-10:], exp_seq3.upper())
idx += 1
def _get_repeat_consensus_record(self):
repeat_consensus_table = self.db.getTable('repeat_consensus')
repeat_consensus_id = self._table_rows['repeat_feature']['repeat_consensus_id']
record = sql.select([repeat_consensus_table],
repeat_consensus_table.c.repeat_consensus_id == repeat_consensus_id)
record = asserted_one(record.execute().fetchall())
self._table_rows['repeat_consensus'] = record
limit_length = lambda x : DisplayString(x, repr_length=10)
attr_column_map = [('Symbol', 'repeat_name', _quoted),
('RepeatClass', 'repeat_class', _quoted),
('RepeatType', 'repeat_type', _quoted),
('Consensus', 'repeat_consensus', limit_length)]
self._populate_cache_from_record(attr_column_map, 'repeat_consensus')
def _get_cpg_island_analysis_id(self):
analysis_description_table = \
self.genome.CoreDb.getTable('analysis_description')
query = sql.select(
[analysis_description_table.c.analysis_id],
analysis_description_table.c.display_label.like('%CpG%'))
record = asserted_one(query.execute())
self._table_rows['analysis_description'] = record
quoted_limited = lambda x: DisplayString(
x, with_quotes=True, num_words=2)
self._populate_cache_from_record(
[('CpGisland', 'analysis_id', quoted_limited)],
'analysis_description')
def _get_seq_region_record(self, CoordName):
# override the _Region class method, since, we take the provided Start
# etc .. attributes
# CoordName comes from seq_region_table.c.name
# matched, by coord_system_id, to default coord system
seq_region_table = self.genome.db.getTable('seq_region')
coord_systems = CoordSystem(core_db=self.genome.CoreDb)
coord_system_ids = [k for k in coord_systems if not isinstance(k, str)]
record = sql.select([seq_region_table],
sql.and_(seq_region_table.c.name == CoordName,
seq_region_table.c.coord_system_id.in_(coord_system_ids)))
record = asserted_one(record.execute().fetchall())
self._table_rows['seq_region'] = record
def test_intron(self):
"""should get correct Intron sequence, regardless of strand"""
# IL2 is on - strand, IL13 is on + strand, both have three introns
IL2_exp_introns = [
(1, 123377358, 123377448, 'gtaagtatat', 'actttcttag'),
(2, 123375008, 123377298, 'gtaagtacaa', 'attattctag'),
(3, 123373017,123374864, 'gtaaggcatt', 'tcttttatag')]
IL13_exp_introns = [
(1, 131994052, 131995109, 'gtgagtgtcg', 'gctcccacag'),
(2, 131995163, 131995415, 'gtaaggacct', 'ctccccacag'),
(3, 131995520, 131995866, 'gtaaggcatc', 'tgtcctgcag')]
for symbol, stable_id, exp_introns in [
('IL2', 'ENST00000226730', IL2_exp_introns),
('IL13', 'ENST00000304506', IL13_exp_introns)]:
gene = asserted_one(self.human.getGenesMatching(Symbol=symbol))
strand = gene.Location.Strand
transcript = asserted_one(
[t for t in gene.Transcripts if t.StableId==stable_id])
introns = transcript.Introns
self.assertEqual(len(introns), len(exp_introns))
idx = 0
for intron in introns:
loc = intron.Location
start, end = loc.Start, loc.End
seq = str(intron.Seq)
exp_rank, exp_start, exp_end, exp_seq5, \
exp_seq3 = exp_introns[idx]
self.assertEqual(loc.Strand, strand)
# test the order using rank
self.assertEqual(intron.Rank, exp_rank)
# test position
self.assertEqual(start, exp_start)
self.assertEqual(end, exp_end)
# test sequence
self.assertEqual(seq[:10], exp_seq5.upper())
self.assertEqual(seq[-10:], exp_seq3.upper())
idx += 1
def __init__(self, genome, db, StableId=None, Symbol=None, Location=None, data=None):
"""constructed by a genome instance"""
super(Gene, self).__init__(genome, db, Location=Location)
if data is None:
args = [dict(StableId=StableId), dict(Symbol=Symbol)][StableId is None]
assert args
data = asserted_one(list(self.genome._get_gene_query(db, **args).execute()))
for name, func in \
[('StableId',self._get_gene_stable_id_record),
('BioType', self._get_gene_record),
('Description', self._get_gene_record),
('Symbol', self._get_xref_record),
('Location', self._get_gene_record)]:
if name == 'Symbol' and 'display_label' not in data.keys(): # For EST
continue
self._table_rows[self._attr_ensembl_table_map[name]] = data
func() # this populates the attributes
def assembly_exception_coordinate(loc):
"""returns a coordinate conversion for one with an assembly exception"""
genome = loc.genome
assemb_except_table = genome.CoreDb.getTable('assembly_exception')
seq_region_table = genome.CoreDb.getTable('seq_region')
query = sql.select([assemb_except_table, seq_region_table.c.name],
sql.and_(
assemb_except_table.c.seq_region_id == \
loc.seq_region_id,
assemb_except_table.c.exc_seq_region_id == \
seq_region_table.c.seq_region_id))
query = location_query(assemb_except_table,
loc.Start, loc.End, query = query)
record = asserted_one(query.execute().fetchall())
s, conv_loc = _get_equivalent_coords(loc, record, "seq_region",
"exc_seq_region", loc.CoordType)
return conv_loc
def assembly_exception_coordinate(loc):
"""returns a coordinate conversion for one with an assembly exception"""
genome = loc.genome
assemb_except_table = genome.CoreDb.getTable('assembly_exception')
seq_region_table = genome.CoreDb.getTable('seq_region')
query = sql.select([assemb_except_table, seq_region_table.c.name],
sql.and_(
assemb_except_table.c.seq_region_id == \
loc.seq_region_id,
assemb_except_table.c.exc_seq_region_id == \
seq_region_table.c.seq_region_id))
query = location_query(assemb_except_table,
loc.Start, loc.End, query = query)
record = asserted_one(query.execute().fetchall())
s, conv_loc = _get_equivalent_coords(loc, record, "seq_region",
"exc_seq_region", loc.CoordType)
return conv_loc
def _make_location(self):
row = self._table_rows[self._attr_ensembl_table_map['Location']]
if row is None:
return
seq_region_id = row['%sid' % self._location_column_prefix]
start = row['%sstart' % self._location_column_prefix]
end = row['%send' % self._location_column_prefix]
strand = row['%sstrand' % self._location_column_prefix]
seq_region_table = self.db.getTable('seq_region')
query = sql.select([seq_region_table.c.name],
seq_region_table.c.seq_region_id == seq_region_id)
result = asserted_one(query.execute().fetchall())
coord_name = result['name']
coord = Coordinate(genome = self.genome, CoordName=coord_name,
Start=start, End=end, Strand=strand,
seq_region_id=seq_region_id,
ensembl_coord=True)
self._cached['Location'] = coord
def _get_coord_type_and_seq_region_id(coord_name, core_db):
seq_region_table = core_db.getTable('seq_region')
rows = sql.select([seq_region_table]).\
where(seq_region_table.c.name == coord_name).execute().fetchall()
species_coord_sys = CoordSystem(species=core_db.db_name.Species,
core_db=core_db)
try:
selected_row = asserted_one(rows)
except ValueError:
selected_row = None
for row in rows:
# not a default_version
if not row['coord_system_id'] in species_coord_sys:
continue
elif not selected_row:
selected_row = row
break
if selected_row is None:
raise ValueError("Ambigous coordinate name: %s" % coord_name)
coord_type = species_coord_sys[selected_row['coord_system_id']].name
return selected_row, coord_type
def _get_coord_type_and_seq_region_id(coord_name, core_db):
seq_region_table = core_db.getTable('seq_region')
rows = sql.select([seq_region_table]).\
where(seq_region_table.c.name == str(coord_name)).execute().fetchall()
species_coord_sys = CoordSystem(species=core_db.db_name.Species,
core_db = core_db)
try:
selected_row = asserted_one(rows)
except ValueError:
selected_row = None
for row in rows:
# not a default_version
if not row['coord_system_id'] in species_coord_sys:
continue
elif not selected_row:
selected_row = row
break
if selected_row is None:
raise ValueError("Ambigous coordinate name: %s" % coord_name)
coord_type = species_coord_sys[selected_row['coord_system_id']].name
return selected_row, coord_type
def _get_simple_features(self, db, klass, target_coord, query_coord,
where_feature):
"""returns feature_type records for the query_coord from the
simple_feature table. The returned coord is referenced to
target_coord. At present, only CpG islands being queried."""
simple_feature_table = db.getTable('simple_feature')
feature_types = ['CpGisland']
feature_type_ids = [
self._feature_type_ids.get(f) for f in feature_types
]
# fix the following
query = sql.select(
[simple_feature_table],
sql.and_(
simple_feature_table.c.analysis_id.in_(feature_type_ids),
simple_feature_table.c.seq_region_id ==
query_coord.seq_region_id))
query = location_query(simple_feature_table,
query_coord.EnsemblStart,
query_coord.EnsemblEnd,
query=query,
where=where_feature)
records = query.execute()
for record in records:
coord = Coordinate(self,
CoordName=query_coord.CoordName,
Start=record['seq_region_start'],
End=record['seq_region_end'],
seq_region_id=record['seq_region_id'],
Strand=record['seq_region_strand'],
ensembl_coord=True)
if query_coord.CoordName != target_coord.CoordName:
coord = asserted_one(
get_coord_conversion(coord, target_coord.CoordType,
self.CoreDb))[1]
# coord = coord.makeRelativeTo(query_coord) #TODO: fix here if query_coord and target_coord have different coordName
# coord = coord.makeRelativeTo(target_coord, False)
yield klass(self, db, Location=coord, Score=record['score'])
def _get_sequence_from_direct_assembly(coord=None, DEBUG=False):
# TODO clean up use of a coord
genome = coord.genome
# no matter what strand user provide, we get the + sequence first
coord.Strand = 1
species = genome.Species
coord_type = CoordSystem(species=species,
core_db=genome.CoreDb,
seq_level=True)
if DEBUG:
print('Created Coordinate:', coord, coord.EnsemblStart,
coord.EnsemblEnd)
print(coord.CoordType, coord_type)
assemblies = get_coord_conversion(coord, coord_type, genome.CoreDb)
if not assemblies:
raise NoItemError('no assembly for %s' % coord)
dna = genome.CoreDb.getTable('dna')
seqs, positions = [], []
for q_loc, t_loc in assemblies:
assert q_loc.Strand == 1
length = len(t_loc)
# get MySQL to do the string slicing via substr function
query = sql.select([
substr(dna.c.sequence, t_loc.EnsemblStart,
length).label('sequence')
], dna.c.seq_region_id == t_loc.seq_region_id)
record = asserted_one(query.execute().fetchall())
seq = record['sequence']
seq = DNA.makeSequence(seq)
if t_loc.Strand == -1:
seq = seq.rc()
seqs.append(str(seq))
positions.append((q_loc.Start, q_loc.End))
sequence = _assemble_seq(seqs, coord.Start, coord.End, positions)
return sequence
def _get_dnafrag_id_for_coord(self, coord):
"""returns the dnafrag_id for the coordnate"""
dnafrag_table = self.ComparaDb.getTable('dnafrag')
genome_db_table = self.ComparaDb.getTable('genome_db')
# column renamed between versions
prefix = coord.genome.Species.lower()
if int(self.Release) > 58:
prefix = _Species.getEnsemblDbPrefix(prefix)
query = sql.select([dnafrag_table.c.dnafrag_id,
dnafrag_table.c.coord_system_name],
sql.and_(dnafrag_table.c.genome_db_id ==\
genome_db_table.c.genome_db_id,
genome_db_table.c.name == prefix,
dnafrag_table.c.name == coord.CoordName))
try:
record = asserted_one(query.execute().fetchall())
dnafrag_id = record['dnafrag_id']
except NoItemError:
raise RuntimeError, 'No DNA fragment identified'
return dnafrag_id
def _get_dnafrag_id_for_coord(self, coord):
"""returns the dnafrag_id for the coordnate"""
dnafrag_table = self.ComparaDb.getTable('dnafrag')
genome_db_table = self.ComparaDb.getTable('genome_db')
# column renamed between versions
prefix = coord.genome.Species.lower()
if int(self.Release) > 58:
prefix = _Species.getEnsemblDbPrefix(prefix)
query = sql.select([dnafrag_table.c.dnafrag_id,
dnafrag_table.c.coord_system_name],
sql.and_(dnafrag_table.c.genome_db_id ==\
genome_db_table.c.genome_db_id,
genome_db_table.c.name == prefix,
dnafrag_table.c.name == coord.CoordName))
try:
record = asserted_one(query.execute().fetchall())
dnafrag_id = record['dnafrag_id']
except NoItemError:
raise RuntimeError, 'No DNA fragment identified'
return dnafrag_id
def _get_sequence_from_direct_assembly(coord=None, DEBUG=False):
# TODO clean up use of a coord
genome = coord.genome
# no matter what strand user provide, we get the + sequence first
coord.Strand = 1
species = genome.Species
coord_type = CoordSystem(species=species,core_db=genome.CoreDb,
seq_level=True)
if DEBUG:
print 'Created Coordinate:',coord,coord.EnsemblStart,coord.EnsemblEnd
print coord.CoordType, coord_type
assemblies = get_coord_conversion(coord, coord_type, genome.CoreDb)
if not assemblies:
raise NoItemError, 'no assembly for %s' % coord
dna = genome.CoreDb.getTable('dna')
seqs, positions = [], []
for q_loc, t_loc in assemblies:
assert q_loc.Strand == 1
length = len(t_loc)
# get MySQL to do the string slicing via substr function
query = sql.select([substr(dna.c.sequence,
t_loc.EnsemblStart,
length).label('sequence')],
dna.c.seq_region_id == t_loc.seq_region_id)
record = asserted_one(query.execute().fetchall())
seq = record['sequence']
seq = DNA.makeSequence(seq)
if t_loc.Strand == -1:
seq = seq.rc()
seqs.append(str(seq))
positions.append((q_loc.Start, q_loc.End))
sequence = _assemble_seq(seqs, coord.Start, coord.End, positions)
return sequence
