mirror of
https://github.com/Priyatham-sai-chand/compara-deep-learning.git
synced 2026-10-05 08:11:34 -07:00
Add files via upload
This commit is contained in:
parent
647b2c9140
commit
855c89df56
1 changed files with 79 additions and 103 deletions
|
|
@ -3,19 +3,14 @@ import requests
|
||||||
import progressbar
|
import progressbar
|
||||||
import sys
|
import sys
|
||||||
|
|
||||||
|
|
||||||
def update_protein(gene_seq,gene):
|
def update_protein(gene_seq,gene):
|
||||||
t=0
|
t=0
|
||||||
while(t!=2):
|
while(t!=2):
|
||||||
try:
|
try:
|
||||||
server = "https://rest.ensembl.org"
|
server = "https://rest.ensembl.org"
|
||||||
ext = "/sequence/id/" + \
|
ext = "/sequence/id/"+str(gene)+"?type=protein;multiple_sequences=1"
|
||||||
str(gene) + "?type=protein;multiple_sequences=1"
|
|
||||||
|
|
||||||
r = requests.get(
|
r = requests.get(server+ext, headers={ "Content-Type" : "application/json"})
|
||||||
server + ext,
|
|
||||||
headers={
|
|
||||||
"Content-Type": "application/json"})
|
|
||||||
|
|
||||||
if not r.ok:
|
if not r.ok:
|
||||||
r.raise_for_status()
|
r.raise_for_status()
|
||||||
|
|
@ -36,16 +31,15 @@ def update_protein(gene_seq, gene):
|
||||||
r=dict(r[maxi])
|
r=dict(r[maxi])
|
||||||
gene_seq[gene]=str(r["seq"])
|
gene_seq[gene]=str(r["seq"])
|
||||||
return
|
return
|
||||||
except BaseException:
|
except :
|
||||||
t+=1
|
t+=1
|
||||||
#print("\nError:",e)
|
#print("\nError:",e)
|
||||||
continue
|
continue
|
||||||
gene_seq[gene]=""
|
gene_seq[gene]=""
|
||||||
|
|
||||||
|
def update_rest_protein(data,fname):
|
||||||
def update_rest_protein(data):
|
|
||||||
gids={}
|
gids={}
|
||||||
with open("processed/not_found.json", "r") as file:
|
with open("processed/not_found_"+fname+".json","r") as file:
|
||||||
gids=dict(json.load(file))
|
gids=dict(json.load(file))
|
||||||
|
|
||||||
gids=list(gids.keys())
|
gids=list(gids.keys())
|
||||||
|
|
@ -54,19 +48,13 @@ def update_rest_protein(data):
|
||||||
|
|
||||||
server = "https://rest.ensembl.org"
|
server = "https://rest.ensembl.org"
|
||||||
ext = "/sequence/id?type=protein"
|
ext = "/sequence/id?type=protein"
|
||||||
headers = {
|
headers={ "Content-Type" : "application/json", "Accept" : "application/json"}
|
||||||
"Content-Type": "application/json",
|
|
||||||
"Accept": "application/json"}
|
|
||||||
|
|
||||||
for i in progressbar.progressbar(range(0,len(gids)-50,50)):
|
for i in progressbar.progressbar(range(0,len(gids)-50,50)):
|
||||||
ids=dict(ids=list(gids[i:i+50]))
|
ids=dict(ids=list(gids[i:i+50]))
|
||||||
while(1):
|
while(1):
|
||||||
try:
|
try:
|
||||||
r = requests.post(
|
r = requests.post(server+ext, headers=headers, data=str(json.dumps(ids)))
|
||||||
server + ext,
|
|
||||||
headers=headers,
|
|
||||||
data=str(
|
|
||||||
json.dumps(ids)))
|
|
||||||
if not r.ok:
|
if not r.ok:
|
||||||
r.raise_for_status()
|
r.raise_for_status()
|
||||||
gs=r.json()
|
gs=r.json()
|
||||||
|
|
@ -83,26 +71,21 @@ def update_rest_protein(data):
|
||||||
for genes in gids:
|
for genes in gids:
|
||||||
try:
|
try:
|
||||||
_=data[genes]
|
_=data[genes]
|
||||||
except BaseException:
|
except:
|
||||||
print(genes)
|
print(genes)
|
||||||
update_protein(data,genes)
|
update_protein(data,genes)
|
||||||
|
|
||||||
print("Gene Sequences Updated Successfully")
|
print("Protein Sequences Updated Successfully")
|
||||||
return data
|
return data
|
||||||
|
|
||||||
|
|
||||||
def update(gene_seq,gene):
|
def update(gene_seq,gene):
|
||||||
t=0
|
t=0
|
||||||
while(t!=2):
|
while(t!=2):
|
||||||
try:
|
try:
|
||||||
server = "https://rest.ensembl.org"
|
server = "https://rest.ensembl.org"
|
||||||
ext = "/sequence/id/" + \
|
ext = "/sequence/id/"+str(gene)+"?type=cds;multiple_sequences=1"
|
||||||
str(gene) + "?type=cds;multiple_sequences=1"
|
|
||||||
|
|
||||||
r = requests.get(
|
r = requests.get(server+ext, headers={ "Content-Type" : "application/json"})
|
||||||
server + ext,
|
|
||||||
headers={
|
|
||||||
"Content-Type": "application/json"})
|
|
||||||
|
|
||||||
if not r.ok:
|
if not r.ok:
|
||||||
r.raise_for_status()
|
r.raise_for_status()
|
||||||
|
|
@ -123,13 +106,12 @@ def update(gene_seq, gene):
|
||||||
r=dict(r[maxi])
|
r=dict(r[maxi])
|
||||||
gene_seq[gene]=str(r["seq"])
|
gene_seq[gene]=str(r["seq"])
|
||||||
return
|
return
|
||||||
except BaseException:
|
except Exception as e:
|
||||||
t+=1
|
t+=1
|
||||||
#print("\nError:",e)
|
#print("\nError:",e)
|
||||||
continue
|
continue
|
||||||
gene_seq[gene]=""
|
gene_seq[gene]=""
|
||||||
|
|
||||||
|
|
||||||
def update_rest(data,fname):
|
def update_rest(data,fname):
|
||||||
gids={}
|
gids={}
|
||||||
with open("processed/not_found_"+fname+".json","r") as file:
|
with open("processed/not_found_"+fname+".json","r") as file:
|
||||||
|
|
@ -141,19 +123,13 @@ def update_rest(data, fname):
|
||||||
|
|
||||||
server = "https://rest.ensembl.org"
|
server = "https://rest.ensembl.org"
|
||||||
ext = "/sequence/id?type=cds"
|
ext = "/sequence/id?type=cds"
|
||||||
headers = {
|
headers={ "Content-Type" : "application/json", "Accept" : "application/json"}
|
||||||
"Content-Type": "application/json",
|
|
||||||
"Accept": "application/json"}
|
|
||||||
|
|
||||||
for i in progressbar.progressbar(range(0,len(gids)-50,50)):
|
for i in progressbar.progressbar(range(0,len(gids)-50,50)):
|
||||||
ids=dict(ids=list(gids[i:i+50]))
|
ids=dict(ids=list(gids[i:i+50]))
|
||||||
while(1):
|
while(1):
|
||||||
try:
|
try:
|
||||||
r = requests.post(
|
r = requests.post(server+ext, headers=headers, data=str(json.dumps(ids)))
|
||||||
server + ext,
|
|
||||||
headers=headers,
|
|
||||||
data=str(
|
|
||||||
json.dumps(ids)))
|
|
||||||
if not r.ok:
|
if not r.ok:
|
||||||
r.raise_for_status()
|
r.raise_for_status()
|
||||||
gs=r.json()
|
gs=r.json()
|
||||||
|
|
@ -170,7 +146,7 @@ def update_rest(data, fname):
|
||||||
for genes in gids:
|
for genes in gids:
|
||||||
try:
|
try:
|
||||||
_=data[genes]
|
_=data[genes]
|
||||||
except BaseException:
|
except:
|
||||||
print(genes)
|
print(genes)
|
||||||
update(data,genes)
|
update(data,genes)
|
||||||
|
|
||||||
|
|
|
||||||
Loading…
Reference in a new issue