def build_protein_maps(accessions):
protein_acc_map = cookbook.DictOfSets()
gene_protein_map = cookbook.DictOfSets()
for id, protein_acc, protein_id in accessions:
protein_acc_map[protein_acc.split('.')[0]].add(protein_id)
gene_protein_map[id].add(protein_id)
return protein_acc_map, gene_protein_map
def get_protein_map():
"""
Query Entrez to get a map from its protein accessions to ids and xrefs
"""
result = ProteinMap(acc_2_id=cookbook.DictOfSets(),
xrefs=cookbook.DictOfSets())
for acc, id, refs in refs_for_mouse_protein_accs():
result.acc_2_id[acc].add(id.acc)
for ref in refs:
result.xrefs[id.acc].add(ref)
return result
def get_protein_accession_map():
"""
Query Entrez to get a map from its protein accession to ids
"""
from Bio.EUtils import HistoryClient
search_term = 'mouse[orgn]'
#search_term = 'MYOD[Gene name] AND mouse[orgn]'
# get a handle to the results
client = HistoryClient.HistoryClient()
results = client.search(db='protein', term=search_term)
results_size = len(results)
dbids = results.dbids.ids
print '# results: %d' % results_size
# download them bit by bit
acc_2_id = cookbook.DictOfSets()
step = 10000
for start in xrange(0, results_size, step):
size = min(step, results_size - start)
end = start + size
results.retstart = start
results.retmax = end
print 'Getting %d->%d' % (start, end)
for id, acc in zip(dbids[start:end],
results.efetch(retmode='text', rettype='acc')):
acc = acc.strip().split('.')[0]
print acc, id
acc_2_id[acc].add(
biopsy.transfac.DbRef.parse_as(
id, biopsy.transfac.db.entrez_protein))
return acc_2_id
def build_mgi_go_map():
"""
@return: A map between MGI identifiers and GO ontologies.
"""
import csv, cookbook
from biopsy import DbRef, db
reader = csv.reader(open(gene_association_filename, "r"), delimiter='\t')
result = cookbook.DictOfSets()
for row in reader:
if row[0].startswith('!'):
continue
result[DbRef.parse_as(row[1], db.mgi)].add(row[4])
return result
def _mgi_ids_from_biomart():
"Goes to biomart to map all mouse genes to MGI ids"
from . import entrez
result = cookbook.DictOfSets()
# build biomart query for swissprot and execute
query = biomart.new_query()
dataset = biomart.add_dataset(query, 'mmusculus_gene_ensembl')
biomart.add_attribute(dataset, 'ensembl_gene_id')
biomart.add_attribute(dataset, 'external_gene_id')
for row in csv.reader(biomart.execute_query(query), delimiter=','):
gene_ref = biopsy.transfac.DbRef.parse_as(row[0],
biopsy.transfac.db.ensembl)
if row[1]:
mgi_acc = row[1]
if mgi_acc in mgi.acc2id():
yield gene_ref, mgi.acc2id()[mgi_acc]
def get_orthologs(species1, species2):
"Returns a dict mapping ensembl genes of species 1 to genes of species 2."
result = cookbook.DictOfSets()
# build biomart query and execute
query = biomart.new_query()
dataset = biomart.add_dataset(query, species[species2].dataset())
biomart.add_attribute(dataset, 'ensembl_gene_id')
biomart.add_attribute(dataset,
'%s_ensembl_gene' % species[species1].short_name)
biomart.print_query(query, 'query.xml')
for row in csv.reader(biomart.execute_query(query), delimiter=','):
if row[1]:
ref_1 = biopsy.transfac.DbRef.parse_as(row[0],
biopsy.transfac.db.ensembl)
ref_2 = biopsy.transfac.DbRef.parse_as(row[1],
biopsy.transfac.db.ensembl)
result[ref_2].add(ref_1)
return result
def proteins_for_species(species):
"Goes to biomart to map all genes in species to Swissprot proteins"
from . import entrez
result = cookbook.DictOfSets()
# build biomart query for swissprot and execute
query = biomart.new_query()
dataset = biomart.add_dataset(query, species)
biomart.add_attribute(dataset, 'ensembl_gene_id')
biomart.add_attribute(dataset, 'ensembl_transcript_id')
biomart.add_attribute(dataset, 'uniprot_swissprot_accession')
for row in csv.reader(biomart.execute_query(query), delimiter=','):
gene_ref = biopsy.transfac.DbRef.parse_as(row[0],
biopsy.transfac.db.ensembl)
transcript_ref = biopsy.transfac.DbRef.parse_as(
row[1], biopsy.transfac.db.ensembl)
if row[2]:
protein_ref = biopsy.transfac.DbRef.parse_as(
row[2], biopsy.transfac.db.swissprot)
result[gene_ref].add((transcript_ref, protein_ref))
# build biomart query for entrez protein and execute
query = biomart.new_query()
dataset = biomart.add_dataset(query, species)
biomart.add_attribute(dataset, 'ensembl_gene_id')
biomart.add_attribute(dataset, 'ensembl_transcript_id')
biomart.add_attribute(dataset, 'protein')
for row in csv.reader(biomart.execute_query(query), delimiter=','):
gene_ref = biopsy.transfac.DbRef.parse_as(row[0],
biopsy.transfac.db.ensembl)
transcript_ref = biopsy.transfac.DbRef.parse_as(
row[1], biopsy.transfac.db.ensembl)
if row[2]:
protein_acc = row[2]
#print "'%s'" % protein_acc
if protein_acc in entrez.mouse_proteins().acc2id:
for protein_id in entrez.mouse_proteins().acc2id[protein_acc]:
ref = T.DbRef(T.db.entrez_protein, "", protein_id)
result[gene_ref].add((transcript_ref, ref))
return result
def get_protein_accession_map_2():
import elementtree.ElementTree as ET
search_term = 'mouse[orgn]'
#search_term = 'MYOD[Gene name] AND mouse[orgn]'
client = HistoryClient.HistoryClient()
results = client.search(db='protein', term=search_term)
results_size = len(results)
print '# results: %d' % results_size
acc_2_ids = cookbook.DictOfSets()
step = 5000
for start in xrange(0, results_size, step):
results.retstart = start
results.retmax = min(step, results_size - start)
#results.retmax = 1000
start = time.time()
summary = results.summary()
print 'Retrieving summary: %f secs' % (time.time() - start)
for entry in summary:
acc = entry.dataitems['Caption'].encode().strip().split('.')[0]
acc_2_ids[acc].add(
biopsy.transfac.DbRef.parse_as(
entry.id.encode(), biopsy.transfac.db.entrez_protein))