def load_gaf(filename, collection, start=0):
print("Loading: %s" % filename)
collection.create_index("DB_Object_ID")
collection.create_index("DB")
collection.create_index("GO_ID")
collection.create_index("Evidence")
collection.create_index("Aspect")
collection.create_index("Date")
n = count_lines(filename, sep=bytes('\n', 'ascii'))
pbar = tqdm(range(n), desc="annotations loaded")
with open(filename, 'r') as handler:
goa_iterator = gafiterator(handler)
for i, data in enumerate(goa_iterator):
if i < start \
or (args.noiea and data['Evidence'] == 'IEA') \
or (args.exp and data['Evidence'] not in exp_codes):
pbar.update(1)
continue
date = datetime.datetime.strptime(data['Date'], "%Y%m%d").date()
json = {
"DB_Object_ID": data['DB_Object_ID'],
"DB_Object_Symbol": data['DB_Object_Symbol'],
"With": data['With'],
"Assigned_By": data['Assigned_By'],
"Annotation_Extension": data['Annotation_Extension'],
"Gene_Product_Form_ID": data['Gene_Product_Form_ID'],
"DB:Reference": data['DB:Reference'],
"GO_ID": data['GO_ID'],
"Qualifier": data['Qualifier'],
"Date": datetime.datetime.fromordinal(date.toordinal()),
"DB": data['DB'],
"created_at": datetime.datetime.utcnow(),
"DB_Object_Name": data['DB_Object_Name'],
"DB_Object_Type": data['DB_Object_Type'],
"Evidence": data['Evidence'],
"Taxon_ID": data['Taxon_ID'],
"Aspect": data['Aspect']
}
collection.update_one({"_id": i}, {'$set': json}, upsert=True)
pbar.update(1)
pbar.close()
print("\nFinished!")
def go_annotations(filepath: str):
"""Handles GO annotation file io.
# Arguments
filepath: str, filepath of GO annotation file
# Returns
gafiterator
# Raises
FileNotFoundError
"""
try:
with open(filepath) as f:
iterator = gafiterator(f)
yield iterator
except FileNotFoundError as err:
raise
def parse_gaf(fn):
if os.path.splitext(fn)[1].lower()[-3:] == '.gz':
open_func = gzip.open
else:
open_func = open
with open_func(fn, 'rb') as f:
it = gafiterator(f)
gene_by_go_term = {}
go_term_by_gene = {}
for rec in it:
gs = rec['DB_Object_Symbol']
go_id = rec['GO_ID']
if go_id not in gene_by_go_term:
gene_by_go_term[go_id] = []
if gs not in go_term_by_gene:
go_term_by_gene[gs] = []
gene_by_go_term[go_id].append(gs)
go_term_by_gene[gs].append(go_id)
return gene_by_go_term, go_term_by_gene
def getAnnotations():
# uses database from uniprot to count the number of annotations each GO term has, and returns a dictionary of GO IDs and their number of annotations
annotDict = {}
for goID in G.nodes():
annotDict[goID] = 0
filename = 'data/goa_uniprot_all.gaf.gz'
with gzip.open(filename, 'rt') as fp:
count = 0
for annotation in gafiterator(fp):
count += 1
percentage = float(count) * 100 / 424606000
if percentage > 5:
print('hi')
count = 0
goID = unicode(annotation['GO_ID'])
try:
annotDict[goID] += 1
except KeyError:
print(goID)
return annotDict
newEnsemblIDs = []
for i in EnsemblIDs:
if i in UP_ID['EnsemblID']:
newEnsemblIDs.append(i)
C_int_UP = list(map(lambda x: ensembl_to_up[x], newEnsemblIDs))
#Enrichment Analysis
go = obo.GODag('/disks/strw13/DBDM/A4_2/go-basic.obo')
with gzip.open('goa_human.gaf.gz', 'rt') as fp:
funcs = {}
for entry in gafiterator(fp):
uniprot_id = entry.pop('DB_Object_ID')
funcs[uniprot_id] = entry
pop = funcs.keys()
assoc = {}
for x in funcs:
if x not in assoc:
assoc[x] = set()
assoc[x].add(str(funcs[x]['GO_ID']))
dictionary = {x: funcs[x]
for x in funcs
if x in C_int_UP}
#handle = "goa_cow.gaf"
#handle = "goa_dog.gaf"
#handle = "goa_pig.gaf"
#handle = "goa_fly.gaf"
#handle = "goa_worm.gaf"
#handle = "goa_yeast.gaf"
#handle = open(handle)
genes = defaultdict(Gene)
reg_targets = defaultdict(int)
annext_cnt = 0
handle = open(gaffile)
for rec in gafiterator(handle):
if rec["Annotation_Extension"] and rec["DB"] == "UniProtKB":
# print rec["Annotation_Extension"]
protid = rec["DB_Object_ID"]
genes[protid].add_annotation(rec)
annext_cnt += 1
print("%d annotations has extensions" % annext_cnt)
dbs = {"UniProtKB": []}
#%% Genes has_regulation_target
for gene_id, gene in genes.items():
# print("gene", gene_id)
import gzip
g = GODag(obo_file='data/go.obo')
"""
'DB': the protein database;
'DB_Object_ID': protein ID;
'Qualifier': annotation qualifier (such as NOT);
'GO_ID': GO term;
'Evidence': evidence code
"""
# filename = <LOCATION OF GAF FILE>
filename = 'data/goa_human.gaf.gz'
with gzip.open(filename, 'rt') as fp:
x ={annotation['DB_Object_ID']:annotation['GO_ID'] for annotation in gafiterator(fp)}
"""
To get the GO entry for a particular protein's accession number:
g[x['O43707']]
and use .name etc, to get terms of that GO entry
"""
f = 'output/result.pkl'
with open(f, 'rb') as file:
df = pickle.load(file)
go_hits = [g[x[y]] if y in x else '' for y in df['uniprot']]
name_hits = [y.name if y!='' else '' for y in go_hits]