def unique_features(self, features, locus, imgtdb_version):
"""
creates GFE from HLA sequence and locus
:param locus: string containing HLA locus.
:param sequence: string containing sequence data.
:return: GFEobject.
"""
unique = []
if self.features:
for feat in features:
rank = feat.rank if not isutr(feat.term) else 1
feat_str = ":".join([
imgtdb_version, locus,
str(rank),
feat.term.upper(), feat.sequence
])
if not feat_str in self.features:
unique.append(feat)
else:
for feat in features:
feat_q = search_feature(feat.term.upper(), feat.rank,
feat.sequence)
seq_features = self.graph.run(feat_q).to_data_frame()
if seq_features.empty:
unique.append(feat)
return unique
def rename_feat(self, f):
if isutr(f):
return f
else:
return "_".join(f.split("-"))
def get_gfe(self, annotation, locus):
"""
creates GFE from a sequence annotation
:param locus: The gene locus
:type locus: ``str``
:param annotation: An sequence annotation object
:type annotation: ``List``
:rtype: ``List``
Returns:
The GFE notation and the associated features in an array
"""
features = []
accessions = {}
for feat in annotation.annotation:
if isinstance(annotation.annotation[feat], DBSeq) \
or isinstance(annotation.annotation[feat], Seq):
seq = str(annotation.annotation[feat])
else:
seq = str(annotation.annotation[feat].seq)
# TODO: Drop this if statement
if isutr(feat):
feat_str = ":".join([locus, str(1), feat, seq])
# If the feature has been loaded or stored
# then use that instead of making a feature request
if self.verbose and self.verbosity > 2:
self.logger.info("Getting accession " + feat_str)
if feat_str in self.all_feats[locus]:
if self.verbose and self.verbosity > 2:
self.logger.info("Feature found " + feat_str)
accession = self.all_feats[locus][feat_str]
feature = Feature(term=feat,
rank=1,
locus=locus,
sequence=seq,
accession=accession)
accessions.update({feat: accession})
features.append(feature)
else:
if self.verbose and self.verbosity > 2:
self.logger.info(self.logname +
"Making FeatureRequest " + feat_str)
# Create FeatureRequest object
request = FeatureRequest(locus=locus,
term=feat,
rank=1,
sequence=seq)
# Attempt to make feature request
try:
feature = self.api.create_feature(body=request)
accessions.update({feat: feature.accession})
features.append(feature)
except ApiException as e:
self.logger.error(
self.logname +
"Exception when calling DefaultApi->create_feature"
+ e)
blank_feat = Feature(term=feat,
rank=1,
locus=locus,
sequence=seq)
accessions.update({feat: 0})
features.append(blank_feat)
# Store new features for quick retrieval if flag passed
if self.store_features:
# Adding new feature to all_feats
self.all_feats[locus].update(
{feat_str: feature.accession})
# Calculating memory size of all_feats
if self.verbose and self.verbosity > 1:
self.logger.info(self.logname +
"Storing new feature " + feat_str)
mem = "{:4.4f}".format(
sys.getsizeof(self.all_feats) / 1000000)
self.logger.info(self.logname +
"Updated * all_feats " + mem +
" MB *")
else:
term, rank = feat.split("_")
feat_str = ":".join([locus, str(rank), term, seq])
# If the feature has been loaded or stored
# then use that instead of making a feature request
if feat_str in self.all_feats[locus]:
if self.verbose and self.verbosity > 2:
self.logger.info(self.logname + "Feature found " +
feat_str)
accession = self.all_feats[locus][feat_str]
feature = Feature(term=term,
rank=rank,
locus=locus,
sequence=seq,
accession=accession)
accessions.update({feat: accession})
features.append(feature)
else:
if self.verbose and self.verbosity > 2:
self.logger.info(self.logname +
"Making FeatureRequest " + feat_str)
# Create FeatureRequest object
request = FeatureRequest(locus=locus,
term=term,
rank=rank,
sequence=seq)
# Attempt to make feature request
try:
feature = self.api.create_feature(body=request)
accessions.update({feat: feature.accession})
features.append(feature)
except ApiException as e:
self.logger.error(
self.logname +
"Exception when calling DefaultApi->create_feature %e"
+ e)
blank_feat = Feature(term=term,
rank=rank,
locus=locus,
sequence=seq)
accessions.update({feat: 0})
features.append(blank_feat)
# Store new features for quick retrieval if flag passed
if self.store_features:
# Adding new feature to all_feats
self.all_feats[locus].update(
{feat_str: feature.accession})
# Calculating memory size of all_feats
if self.verbose and self.verbosity > 1:
self.logger.info(self.logname +
"Storing new feature " + feat_str)
mem = "{:4.4f}".format(
sys.getsizeof(self.all_feats) / 1000000)
self.logger.info(self.logname +
"Updated * all_feats " + mem +
" MB *")
# Creating GFE
gfe = self._make_gfe(accessions, locus)
if self.verbose:
self.logger.info("GFE = " + gfe)
return features, gfe
def test_001_isutr(self):
self.assertTrue(isutr('three_prime_UTR'))
self.assertTrue(isutr('five_prime_UTR'))
self.assertFalse(isutr('exon-3'))
pass
