P0c6X7.py

import pandas as pd
import numpy as np
import urllib
import os 
import requests
import matplotlib.pyplot as plt
from sklearn.metrics import r2_score
import seaborn as seabornInstance 
from sklearn.model_selection import train_test_split 
from sklearn.linear_model import LinearRegression
from sklearn import metrics

import time





link1 = "https://pubchem.ncbi.nlm.nih.gov/sdq/cgi2rcgi.cgi?infmt=json&outfmt=csv&query={%22download%22:%22*%22,%22collection%22:%22bioactivity%22,%22where%22:{%22ands%22:[{%22protacxn%22:%22notnull%22},{%22cid%22:%22notnull%22},{%22repacxn%22:%22P0C6X7%22}]},%22order%22:[%22activity,asc%22],%22start%22:1,%22limit%22:10000000,%22downloadfilename%22:%22PROTACXN_P0C6X7_bioactivity_protein%22}"





for i in range(0,2):
    try:
        os.remove('downloaded1.csv')
        
        break
    except Exception as e:
        
        break
    else:
        break





for i in range(2):
    try:
        req = requests.get(link1)
        url_content = req.content
        csv_file = open('downloaded1.csv', 'wb')
        csv_file.write(url_content)
        csv_file.close()
        print("Completed the Request")
        break
    
     
      
    except Exception as a:
        print(str(a)+" is the error , Trying {} time".format(i))
        continue
    else:
        break
else:
    print("something Wrong , Try running Again [refer error code for more]")
            





data = pd.read_csv("downloaded1.csv",error_bad_lines=False)

data_df = pd.DataFrame(data)

new_data = data[['cid','acvalue']]


new_data = new_data.dropna()






except_val=0





cid_value = new_data['cid'].to_list()

PIC50_value = (-np.log10(new_data['acvalue']*10**-6)).to_list()





PIC50_value = pd.DataFrame(PIC50_value,columns = ["y"])





y_data = PIC50_value











link = "https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/cid/new_link_me/property/MolecularWeight,HeavyAtomCount,XLOGP,Complexity,HBondAcceptorCount,MonoisotopicMass,RotatableBondCount,TPSA/CSV"
link_fixed = "https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/cid/new_link_me/property/MolecularWeight,HeavyAtomCount,XLOGP,Complexity,HBondAcceptorCount,MonoisotopicMass,RotatableBondCount,TPSA/CSV"
sub_link = "https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/fastsimilarity_2d/cid/replaceme/cids/TXT"
sub_link_fixed = "https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/fastsimilarity_2d/cid/replaceme/cids/TXT"

counter = 1





final_data_frame = pd.DataFrame()









x_data = pd.DataFrame()


cid_main_counter=0 








for x in cid_value:
    for iter_x in range(1000):
        try:    
            link = link_fixed
            link = link.replace("new_link_me",str(x))
            data1 = pd.read_csv(link)
            data1 = pd.DataFrame(data1)
            x_data = x_data.append(data1)
            data1 = 0
        except Exception as e:
            print("Exception Encountered as {} .! , Trying again !Iteration : {}".format(str(e),str(iter_x)))
            time.sleep(5)
            continue
        else:
            break
    
    else:
        print("Something Wrong with the Trials ! Restart The Algorithm!")





x_data_saved = x_data





x_data = x_data_saved











n_cid=[]






for i in cid_value:
    max_tries = 10
    
    for iter_ in range(max_tries):
        try:
            
            


            link = link_fixed
            link = link.replace("new_link_me",str(i))
            f_data = pd.read_csv(link) 
            
            sub_link = sub_link_fixed
            sub_link = sub_link.replace("replaceme",str(i))
            res = urllib.request.urlopen(sub_link)
            data_sub = res.read()
            data_sub = str(data_sub)
            data_sub = data_sub.replace("\\n",",") ; data_sub = data_sub.replace('b'," ")
            data_sub = data_sub.replace("'",""); data_sub = data_sub.replace(" ","")
            n_count_loop = 0 
            
            
            for j in data_sub.split(","):
                n_count_loop+=1
                if(len(j)>1):
                    if(n_count_loop<=15):
                        n_cid.append(j)
                    else:
                        break
                else:
                    pass
             
        except:
            if(except_val>=15):
                break
            else:
                print("Re-Trying")
                except_val+=1
                
                continue
        else:
            break
    






k_count = 0
t_count = 0
phase_count=0






f_data = pd.DataFrame()








for k in n_cid:
    k_count+=1
    t_count+=1
    phase_count+=1
    for sub_iter in range(10):
            try:
                if(t_count>=50):
                    
                    time.sleep(15)
                    t_count=0
                    
                
            
                link = link_fixed
                link = link.replace("new_link_me",str(k))
                
                f_data_df = pd.read_csv(link)
                f_data = f_data.append(f_data_df)
                
                break

            except Exception as e:
                print(str(e) + "Encountered , Please Wait :+ "+str(k_count))
                time.sleep(20)
                continue

    final_data_frame = final_data_frame.append(f_data)
    cid_main_counter+=1
    final_data_frame.to_csv("fdf1.csv")

    print("Data Fetching Continued : " + str(cid_main_counter))
     
    









final_data_frame.drop_duplicates(inplace=True)









x_data = pd.DataFrame(x_data)






y_data = pd.DataFrame(y_data)






train_file = pd.DataFrame()






x_data = x_data.astype("float64")
y_data = y_data.astype("float64")







new = y_data['y'].to_list()





train_file  = x_data
train_file['y'] = new





train_file
cid_reg_list=train_file['CID'].to_list()
train_file.drop('CID',axis=1,inplace =True)











for x in train_file.isnull().any():
    if x == True:
        train_file = train_file.fillna(method='ffill')






x = list(train_file.columns)
x = x[:-1]

x_ = train_file[x].values
y_ = train_file['y'].values

















x_ = x_.astype("float64")
y_ = y_.astype("float64")




X_train, X_test, y_train, y_test = train_test_split(x_, y_, test_size=0.1, random_state=0)





print("Regression Starts")





regressor = LinearRegression()  
regressor.fit(X_train, y_train)






coef_dict = dict(zip(x, regressor.coef_))

y_pred = regressor.predict(X_test)





main_r2 = r2_score(y_test, y_pred, multioutput='uniform_average')
max_ = main_r2 



from itertools import combinations
comb_list = [[]]


def sub(arr,r):
    global comb_list
    for i in r:
        comb = list(combinations(arr,i))
        comb_list.append(comb)
    return comb_list











newone =0

newone = sub(x , [2,3,4,5,6])
del newone[0]


coef_dict  = 0
loop_index = 0
coef_dict = [[]]
r2_score_new =[]
max_r2 = []
index_r2 =[]






for i in range(0,len(newone)):
    
    index_r=0
    for combi_ in newone[loop_index]:
        
        
        features = list(combi_)
        features_=train_file[features].values
        output_=train_file['y'].values
        X_train, X_test, y_train, y_test = train_test_split(features_, output_, test_size=0.1, random_state=0)
        regressor = LinearRegression()  
        regressor.fit(X_train, y_train)
        y_pred = regressor.predict(X_test)
        r2_score_  = r2_score(y_test, y_pred, multioutput='uniform_average')
        r2_score_new.append(r2_score_)
    loop_index+=1
        
    
    max_r2.append(max(r2_score_new))
    index_r = r2_score_new.index(max(r2_score_new))
    index_r2.append(index_r)
    r2_score_new =[]






sec_index = max_r2.index(max(max_r2))



fir_index = index_r2[sec_index]






if main_r2 > max(max_r2):
    r2_features = x
else:
    
    features  = newone[sec_index][fir_index]
maxi_r2 = max(max_r2)











reg_max_r2 = max(max_r2)









for x in train_file.isnull().any():
    if x == True:
        train_file = train_file.fillna(method='ffill')











features  = list(features)





x_trained =train_file[features].values











y_trained = train_file['y'].values











X_train, X_test, y_train, y_test = train_test_split(x_trained, y_trained, test_size=0.1, random_state=0)





regressor = LinearRegression()  
regressor.fit(X_train, y_train)





coef_dict = dict(zip(features, regressor.coef_))
coef_dict = pd.DataFrame(coef_dict , index = [0])





y_pred = regressor.predict(X_test)




saved_features = features

final_data_frame

final_data_frameF1 = final_data_frame[final_data_frame['MolecularWeight'] <= 500]
final_data_frameF2 = final_data_frameF1[final_data_frameF1['XLogP'] <=5.6]

final_data_frame = final_data_frameF2

pred_data_cid = final_data_frame["CID"].to_list()





features.append('CID')





final_data_frame1 = final_data_frame[features]





for x in final_data_frame1.isnull().any():
    if x == True:
        final_data_frame1 = final_data_frame1.fillna(method='ffill')





final_data_frame1.drop(columns='CID',inplace=True)


final_pred = regressor.predict(final_data_frame1)





final_pred = list(final_pred)





final_data_frame1['CID'] = pred_data_cid





final_data_frame1['y_'] = final_pred



saved_final_data_frame1 = final_data_frame1





sorted_final_df = final_data_frame1.sort_values('y_',ascending =0)





final_larg= sorted_final_df.head(30)




final_larg = final_larg[['CID','y_']]





top_cid = final_larg["CID"].to_list()





for i in range(0,2):
    try:
        os.remove('downloaded1.csv')
        
        break
    except Exception as e:
        
        break
    else:
        break
for i in range(0,2):
    try:
        os.remove('fdf1.csv')
        
        break
    except Exception as e:
        
        break
    else:
        break





print("The Top 30 Drug Leads Which are identified with PubChem cid's are: ")
itter_count = 0
for itter in top_cid:
    itter_count+=1
    
    print(str(itter_count)+" : "+str(itter))


import urllib.request
import os
l=[]
for itter in top_cid:
    l.append(itter)
for i in l:
    url = 'https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/cid/{}/SDF'.format(i)
    urllib.request.urlretrieve(url, 'Lig.sdf')
    
    
  
    os.system('babel -isdf Lig.sdf -opdbqt ligand.pdbqt')
    
    
    os.system('vina --config configCLpro.txt')

  
    os.system('vina --config configPLpro.txt')
    
    print(url)
    
    input("Press any key to continue")