def _retrieve_reference(adaptor, primary_id):
# XXX dbxref_qualifier_value
refs = adaptor.execute_and_fetchall(
"SELECT start_pos, end_pos, "
" location, title, authors,"
" dbname, accession"
" FROM bioentry_reference"
" JOIN reference USING (reference_id)"
" LEFT JOIN dbxref USING (dbxref_id)"
" WHERE bioentry_id = %s"
" ORDER BY rank", (primary_id,))
references = []
for start, end, location, title, authors, dbname, accession in refs:
reference = SeqFeature.Reference()
# If the start/end are missing, reference.location is an empty list
if (start is not None) or (end is not None):
if start is not None:
start -= 1 # python counting
reference.location = [SeqFeature.FeatureLocation(start, end)]
# Don't replace the default "" with None.
if authors:
reference.authors = authors
if title:
reference.title = title
reference.journal = location
if dbname == 'PUBMED':
reference.pubmed_id = accession
elif dbname == 'MEDLINE':
reference.medline_id = accession
references.append(reference)
if references:
return {'references': references}
else:
return {}
def _parse_reference(element):
reference = SeqFeature.Reference()
authors = []
scopes = []
tissues = []
journal_name = ''
pub_type = ''
pub_date = ''
for ref_element in element:
if ref_element.tag == NS + 'citation':
pub_type = ref_element.attrib['type']
if pub_type == 'submission':
pub_type += ' to the ' + ref_element.attrib['db']
if 'name' in ref_element.attrib:
journal_name = ref_element.attrib['name']
pub_date = ref_element.attrib.get('date', '')
j_volume = ref_element.attrib.get('volume', '')
j_first = ref_element.attrib.get('first', '')
j_last = ref_element.attrib.get('last', '')
for cit_element in ref_element:
if cit_element.tag == NS + 'title':
reference.title = cit_element.text
elif cit_element.tag == NS + 'authorList':
for person_element in cit_element:
authors.append(person_element.attrib['name'])
elif cit_element.tag == NS + 'dbReference':
self.ParsedSeqRecord.dbxrefs.append(cit_element.attrib['type']
+ ':' + cit_element.attrib['id'])
if cit_element.attrib['type'] == 'PubMed':
reference.pubmed_id = cit_element.attrib['id']
elif ref_element.attrib['type'] == 'MEDLINE':
reference.medline_id = cit_element.attrib['id']
elif ref_element.tag == NS + 'scope':
scopes.append(ref_element.text)
elif ref_element.tag == NS + 'source':
for source_element in ref_element:
if source_element.tag == NS + 'tissue':
tissues.append(source_element.text)
if scopes:
scopes_str = 'Scope: ' + ', '.join(scopes)
else:
scopes_str = ''
if tissues:
tissues_str = 'Tissue: ' + ', '.join(tissues)
else:
tissues_str = ''
# locations cannot be parsed since they are actually written in
# free text inside scopes so all the references are put in the
# annotation.
reference.location = []
reference.authors = ', '.join(authors)
if journal_name:
if pub_date and j_volume and j_first and j_last:
reference.journal = REFERENCE_JOURNAL % dict(name=journal_name,
volume=j_volume, first=j_first, last=j_last, pub_date=pub_date)
else:
reference.journal = journal_name
reference.comment = ' | '.join((pub_type, pub_date, scopes_str, tissues_str))
append_to_annotations('references', reference)
def to_ref(self):
r = SeqFeature.Reference()
r.title = self.title
r.authors = self.authors
r.journal = self.journal
r.medline_id = self.medline_id
r.pubmed_id = self.pubmed_id
return r
def _format_publications(self):
references = []
for pub in self.genome_object.get('publications', []):
if len(pub) != 7:
log('Skipping unparseable publication {}'.format(pub))
ref = SeqFeature.Reference()
if pub[0]:
ref.pubmed_id = str(pub[0])
ref.title = pub[2]
ref.authors = pub[5]
ref.journal = pub[6]
references.append(ref)
return references
def _parse_reference(element):
reference = SeqFeature.Reference()
authors = []
scopes = []
tissues = []
journal_name = ""
pub_type = ""
pub_date = ""
for ref_element in element:
if ref_element.tag == NS + "citation":
pub_type = ref_element.attrib["type"]
if pub_type == "submission":
pub_type += " to the " + ref_element.attrib["db"]
if "name" in ref_element.attrib:
journal_name = ref_element.attrib["name"]
pub_date = ref_element.attrib.get("date", "")
j_volume = ref_element.attrib.get("volume", "")
j_first = ref_element.attrib.get("first", "")
j_last = ref_element.attrib.get("last", "")
for cit_element in ref_element:
if cit_element.tag == NS + "title":
reference.title = cit_element.text
elif cit_element.tag == NS + "authorList":
for person_element in cit_element:
authors.append(person_element.attrib["name"])
elif cit_element.tag == NS + "dbReference":
self.ParsedSeqRecord.dbxrefs.append(
cit_element.attrib["type"]
+ ":"
+ cit_element.attrib["id"]
)
if cit_element.attrib["type"] == "PubMed":
reference.pubmed_id = cit_element.attrib["id"]
elif ref_element.attrib["type"] == "MEDLINE":
reference.medline_id = cit_element.attrib["id"]
elif ref_element.tag == NS + "scope":
scopes.append(ref_element.text)
elif ref_element.tag == NS + "source":
for source_element in ref_element:
if source_element.tag == NS + "tissue":
tissues.append(source_element.text)
if scopes:
scopes_str = "Scope: " + ", ".join(scopes)
else:
scopes_str = ""
if tissues:
tissues_str = "Tissue: " + ", ".join(tissues)
else:
tissues_str = ""
# locations cannot be parsed since they are actually written in
# free text inside scopes so all the references are put in the
# annotation.
reference.location = []
reference.authors = ", ".join(authors)
if journal_name:
if pub_date and j_volume and j_first and j_last:
reference.journal = REFERENCE_JOURNAL % {
"name": journal_name,
"volume": j_volume,
"first": j_first,
"last": j_last,
"pub_date": pub_date,
}
else:
reference.journal = journal_name
reference.comment = " | ".join(
(pub_type, pub_date, scopes_str, tissues_str)
)
append_to_annotations("references", reference)
def SwissIterator(handle):
"""Breaks up a Swiss-Prot/UniProt file into SeqRecord objects.
Every section from the ID line to the terminating // becomes
a single SeqRecord with associated annotation and features.
This parser is for the flat file "swiss" format as used by:
- Swiss-Prot aka SwissProt
- TrEMBL
- UniProtKB aka UniProt Knowledgebase
For consistency with BioPerl and EMBOSS we call this the "swiss"
format. See also the SeqIO support for "uniprot-xml" format.
"""
swiss_records = SwissProt.parse(handle)
for swiss_record in swiss_records:
# Convert the SwissProt record to a SeqRecord
seq = Seq.Seq(swiss_record.sequence, Alphabet.generic_protein)
record = SeqRecord.SeqRecord(
seq,
id=swiss_record.accessions[0],
name=swiss_record.entry_name,
description=swiss_record.description,
features=[_make_seqfeature(*f) for f in swiss_record.features],
)
record.description = swiss_record.description
for cross_reference in swiss_record.cross_references:
if len(cross_reference) < 2:
continue
database, accession = cross_reference[:2]
dbxref = "%s:%s" % (database, accession)
if dbxref not in record.dbxrefs:
record.dbxrefs.append(dbxref)
annotations = record.annotations
annotations['accessions'] = swiss_record.accessions
if swiss_record.created:
annotations['date'] = swiss_record.created[0]
if swiss_record.sequence_update:
annotations[
'date_last_sequence_update'] = swiss_record.sequence_update[0]
if swiss_record.annotation_update:
annotations[
'date_last_annotation_update'] = swiss_record.annotation_update[
0]
if swiss_record.gene_name:
annotations['gene_name'] = swiss_record.gene_name
annotations['organism'] = swiss_record.organism.rstrip(".")
annotations['taxonomy'] = swiss_record.organism_classification
annotations['ncbi_taxid'] = swiss_record.taxonomy_id
if swiss_record.host_organism:
annotations['organism_host'] = swiss_record.host_organism
if swiss_record.host_taxonomy_id:
annotations['host_ncbi_taxid'] = swiss_record.host_taxonomy_id
if swiss_record.comments:
annotations['comment'] = "\n".join(swiss_record.comments)
if swiss_record.references:
annotations['references'] = []
for reference in swiss_record.references:
feature = SeqFeature.Reference()
feature.comment = " ".join("%s=%s;" % k_v
for k_v in reference.comments)
for key, value in reference.references:
if key == 'PubMed':
feature.pubmed_id = value
elif key == 'MEDLINE':
feature.medline_id = value
elif key == 'DOI':
pass
elif key == 'AGRICOLA':
pass
else:
raise ValueError("Unknown key %s found in references" %
key)
feature.authors = reference.authors
feature.title = reference.title
feature.journal = reference.location
annotations['references'].append(feature)
if swiss_record.keywords:
record.annotations['keywords'] = swiss_record.keywords
yield record
def SwissIterator(handle):
"""Break up a Swiss-Prot/UniProt file into SeqRecord objects.
Every section from the ID line to the terminating // becomes
a single SeqRecord with associated annotation and features.
This parser is for the flat file "swiss" format as used by:
- Swiss-Prot aka SwissProt
- TrEMBL
- UniProtKB aka UniProt Knowledgebase
For consistency with BioPerl and EMBOSS we call this the "swiss"
format. See also the SeqIO support for "uniprot-xml" format.
Rather than calling it directly, you are expected to use this
parser via Bio.SeqIO.parse(..., format="swiss") instead.
"""
with as_handle(handle, "rU") as handle:
swiss_records = SwissProt.parse(handle)
for swiss_record in swiss_records:
# Convert the SwissProt record to a SeqRecord
seq = Seq.Seq(swiss_record.sequence, Alphabet.generic_protein)
record = SeqRecord.SeqRecord(
seq,
id=swiss_record.accessions[0],
name=swiss_record.entry_name,
description=swiss_record.description,
features=[_make_seqfeature(*f) for f in swiss_record.features],
)
record.description = swiss_record.description
for cross_reference in swiss_record.cross_references:
if len(cross_reference) < 2:
continue
database, accession = cross_reference[:2]
dbxref = "%s:%s" % (database, accession)
if dbxref not in record.dbxrefs:
record.dbxrefs.append(dbxref)
annotations = record.annotations
annotations["accessions"] = swiss_record.accessions
if swiss_record.protein_existence:
annotations[
"protein_existence"] = swiss_record.protein_existence
if swiss_record.created:
annotations["date"] = swiss_record.created[0]
annotations["sequence_version"] = swiss_record.created[1]
if swiss_record.sequence_update:
annotations[
"date_last_sequence_update"] = swiss_record.sequence_update[
0]
annotations["sequence_version"] = swiss_record.sequence_update[
1]
if swiss_record.annotation_update:
annotations[
"date_last_annotation_update"] = swiss_record.annotation_update[
0]
annotations["entry_version"] = swiss_record.annotation_update[
1]
if swiss_record.gene_name:
annotations["gene_name"] = swiss_record.gene_name
annotations["organism"] = swiss_record.organism.rstrip(".")
annotations["taxonomy"] = swiss_record.organism_classification
annotations["ncbi_taxid"] = swiss_record.taxonomy_id
if swiss_record.host_organism:
annotations["organism_host"] = swiss_record.host_organism
if swiss_record.host_taxonomy_id:
annotations["host_ncbi_taxid"] = swiss_record.host_taxonomy_id
if swiss_record.comments:
annotations["comment"] = "\n".join(swiss_record.comments)
if swiss_record.references:
annotations["references"] = []
for reference in swiss_record.references:
feature = SeqFeature.Reference()
feature.comment = " ".join("%s=%s;" % k_v
for k_v in reference.comments)
for key, value in reference.references:
if key == "PubMed":
feature.pubmed_id = value
elif key == "MEDLINE":
feature.medline_id = value
elif key == "DOI":
pass
elif key == "AGRICOLA":
pass
else:
raise ValueError(
"Unknown key %s found in references" % key)
feature.authors = reference.authors
feature.title = reference.title
feature.journal = reference.location
annotations["references"].append(feature)
if swiss_record.keywords:
record.annotations["keywords"] = swiss_record.keywords
yield record
def project_to_genbank(filename, project, allblocks, construct_id=None):
if construct_id is not None:
blocks = [construct_id]
else:
blocks = project["components"]
seq_obj_lst = []
# For each of the construct in the project
for block_id in blocks:
block = [b for b in allblocks if b["id"] == block_id][0]
if not block:
continue
# Grab the original ID that came from genbank before if available, otherwise the GD Name as the name
if "genbank" in block["metadata"] and "id" in block["metadata"][
"genbank"]:
genbank_id = block["metadata"]["genbank"]["id"]
elif "genbank" in block["metadata"] and "name" in block["metadata"][
"genbank"]:
genbank_id = block["metadata"]["genbank"]["name"]
else:
genbank_id = "GC_DNA"
sequence = build_sequence(block, allblocks)
seq_obj = SeqIO.SeqRecord(
Seq.Seq(sequence, Seq.Alphabet.DNAAlphabet()), genbank_id)
# Create a 'source' feature
sf = SeqFeature.SeqFeature()
sf.type = "source"
sf.location = SeqFeature.FeatureLocation(0, len(seq_obj.seq))
add_GC_info(sf, block, allblocks)
if "genbank" in block["metadata"]:
# Set up all the annotations in the genbank record. These came originally from genbank.
if "annotations" in block["metadata"]["genbank"]:
for annot_key, annot_value in block["metadata"]["genbank"][
"annotations"].iteritems():
seq_obj.annotations[annot_key] = annot_value
# Set up all the references in the genbank record. These came originally from genbank.
if "references" in block["metadata"]["genbank"]:
for ref in block["metadata"]["genbank"]["references"]:
genbank_ref = SeqFeature.Reference()
genbank_ref.authors = ref['authors']
genbank_ref.comment = ref['comment']
genbank_ref.consrtm = ref['consrtm']
genbank_ref.journal = ref['journal']
genbank_ref.medline_id = ref['medline_id']
genbank_ref.pubmed_id = ref['pubmed_id']
genbank_ref.title = ref['title']
if "references" not in seq_obj.annotations:
seq_obj.annotations["references"] = []
seq_obj.annotations["references"].append(genbank_ref)
# Add the original annotations to the source feature
if "feature_annotations" in block["metadata"]["genbank"]:
for annot_key, annot_value in block["metadata"]["genbank"][
"feature_annotations"].iteritems():
sf.qualifiers[annot_key] = annot_value
seq_obj.features.append(sf)
if "description" in block["metadata"]:
seq_obj.description = block["metadata"]["description"]
if "genbank" in block["metadata"] and "name" in block["metadata"][
"genbank"]:
seq_obj.name = block["metadata"]["genbank"]["name"]
elif "name" in block["metadata"]:
seq_obj.name = block["metadata"]["name"].replace(" ", "")[:5]
else:
seq_obj.name = "GC_DNA"
convert_annotations(block, seq_obj, 0)
# Add a block for each of the features, recursively
start = 0
for child_id in block['components']:
child_block = [b for b in allblocks if b["id"] == child_id][0]
start = add_features(child_block, allblocks, seq_obj, start)
seq_obj_lst.append(seq_obj)
SeqIO.write(seq_obj_lst, open(filename, "w"), "genbank")
