def preprocess_som_patient_data(self, patients):
# get the dictionary of gene id mappers
uni2ent, ent2uni = uniprot_mapper.json_to_dict()
res = []
num_empty = 0
for pat_id, ent_ids in patients.items():
# uni_ids = [uid for eid in ent_ids if eid in ent2uni for uid in ent2uni[eid]]
uni_ids = [
uid for eid in ent_ids if eid in ent2uni
for uid in ent2uni[eid]
]
# if there are any matches map them
'''
if len(uni_ids) > 0: res.append({
'pat_id': pat_id,
'mutated_nodes': uni_ids,
})
else: num_empty += 1
'''
res.append({
'pat_id': pat_id,
'mutated_nodes': uni_ids,
})
log('removed patients:', num_empty)
return res
def preprocess_seq_patient_data(self, GE, all_ent_ids):
# get the dictionary of gene id mappers
uni2ent, ent2uni = uniprot_mapper.json_to_dict()
found_ent_ids = [eid in ent2uni for eid in all_ent_ids]
ent_ids = np.array([eid for eid in all_ent_ids if eid in ent2uni])
uni_ids = np.array([ent2uni[eid] for eid in ent_ids])
log('uni_ids:', len(uni_ids))
log('miss_ent_ids:', len(all_ent_ids) - sum(found_ent_ids))
# prune genes whose uniprot id is not found
GE = GE[found_ent_ids]
return GE, uni_ids
def patient_entrez_to_uniprot():
list_of_gene_patient = []
with open(KIRC_PATH, 'r') as csvfile:
csv_reader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(csv_reader)
for row in csv_reader:
if int(row[1]) != 0:
list_of_gene_patient.append(list(row))
u_map = []
e_map = []
with open(UNIPROT_ENTREZ_MAP_FPATH, 'r') as csvfile:
csv_reader = csv.reader(csvfile, delimiter='\t', quotechar='|')
next(csv_reader) # skip header
for row in csv_reader:
u_map.append(row[0])
e_map.append(row[1])
uni_to_entrez, entrez_to_uni = um.json_to_dict()
patient_uniprot_list = []
for row in list_of_gene_patient:
uni_prot = []
patient = row[2]
try:
uni_prot.append(entrez_to_uni[row[1]])
except:
if row[2] in e_map:
uni_prot.append(u_map[e_map.index(row[2])])
else:
print("none")
if len(uni_prot) != 0:
patient_uniprot_list.append([patient, uni_prot])
return patient_uniprot_list
