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
|
|
@ -3,176 +3,152 @@ 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()
|
||||||
sys.exit()
|
sys.exit()
|
||||||
r = r.json()
|
r=r.json()
|
||||||
if len(r) == 1:
|
if len(r)==1:
|
||||||
r = dict(r[0])
|
r=dict(r[0])
|
||||||
gene_seq[gene] = str(r["seq"])
|
gene_seq[gene]=str(r["seq"])
|
||||||
return
|
return
|
||||||
else:
|
else:
|
||||||
maxi = 0
|
maxi=0
|
||||||
maxlen = 0
|
maxlen=0
|
||||||
for i in range(len(r)):
|
for i in range(len(r)):
|
||||||
m = r[i]
|
m=r[i]
|
||||||
m = dict(m)
|
m=dict(m)
|
||||||
if len(m["seq"]) > maxlen:
|
if len(m["seq"])>maxlen:
|
||||||
maxi = i
|
maxi=i
|
||||||
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):
|
||||||
|
gids={}
|
||||||
|
with open("processed/not_found_"+fname+".json","r") as file:
|
||||||
|
gids=dict(json.load(file))
|
||||||
|
|
||||||
def update_rest_protein(data):
|
gids=list(gids.keys())
|
||||||
gids = {}
|
|
||||||
with open("processed/not_found.json", "r") as file:
|
|
||||||
gids = dict(json.load(file))
|
|
||||||
|
|
||||||
gids = list(gids.keys())
|
geneseq={}
|
||||||
|
|
||||||
geneseq = {}
|
|
||||||
|
|
||||||
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()
|
||||||
tgs = {}
|
tgs={}
|
||||||
for g in gs:
|
for g in gs:
|
||||||
tgs[g["query"]] = g["seq"]
|
tgs[g["query"]]=g["seq"]
|
||||||
geneseq.update(tgs)
|
geneseq.update(tgs)
|
||||||
break
|
break
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print("Error:", e)
|
print("Error:",e)
|
||||||
continue
|
continue
|
||||||
|
|
||||||
data.update(geneseq)
|
data.update(geneseq)
|
||||||
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()
|
||||||
sys.exit()
|
sys.exit()
|
||||||
r = r.json()
|
r=r.json()
|
||||||
if len(r) == 1:
|
if len(r)==1:
|
||||||
r = dict(r[0])
|
r=dict(r[0])
|
||||||
gene_seq[gene] = str(r["seq"])
|
gene_seq[gene]=str(r["seq"])
|
||||||
return
|
return
|
||||||
else:
|
else:
|
||||||
maxi = 0
|
maxi=0
|
||||||
maxlen = 0
|
maxlen=0
|
||||||
for i in range(len(r)):
|
for i in range(len(r)):
|
||||||
m = r[i]
|
m=r[i]
|
||||||
m = dict(m)
|
m=dict(m)
|
||||||
if len(m["seq"]) > maxlen:
|
if len(m["seq"])>maxlen:
|
||||||
maxi = i
|
maxi=i
|
||||||
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):
|
||||||
|
gids={}
|
||||||
|
with open("processed/not_found_"+fname+".json","r") as file:
|
||||||
|
gids=dict(json.load(file))
|
||||||
|
|
||||||
def update_rest(data, fname):
|
gids=list(gids.keys())
|
||||||
gids = {}
|
|
||||||
with open("processed/not_found_" + fname + ".json", "r") as file:
|
|
||||||
gids = dict(json.load(file))
|
|
||||||
|
|
||||||
gids = list(gids.keys())
|
geneseq={}
|
||||||
|
|
||||||
geneseq = {}
|
|
||||||
|
|
||||||
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()
|
||||||
tgs = {}
|
tgs={}
|
||||||
for g in gs:
|
for g in gs:
|
||||||
tgs[g["query"]] = g["seq"]
|
tgs[g["query"]]=g["seq"]
|
||||||
geneseq.update(tgs)
|
geneseq.update(tgs)
|
||||||
break
|
break
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print("Error:", e)
|
print("Error:",e)
|
||||||
continue
|
continue
|
||||||
|
|
||||||
data.update(geneseq)
|
data.update(geneseq)
|
||||||
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)
|
||||||
|
|
||||||
print("Gene Sequences Updated Successfully")
|
print("Gene Sequences Updated Successfully")
|
||||||
return data
|
return data
|
||||||
Loading…
Reference in a new issue