结果数据调整并使用knime计算inchikey

首先我的数据结构是csv格式,columns为化合物id.值为化合物所对应的uniprots:
现在想把格式转化为更方便后续处理的格式:

上代码:

import numpy as np
import pandas as pd
import os
import re


def iter_files(path):
# get all file_path under one folder
    file_names =[]
    for root,dirs,files in os.walk(path):
        for file in files:
            file_name = os.path.join(root,file)
            file_names.append(file_name)

    return file_names  
    
def data_reconsitution(path):
# data reconsititution without Smiles                
    files = iter_files(path)
    for file in files:
        if re.findall('\d+_\d+.csv',file):            
            
            df = pd.read_csv(file).iloc[:,1:]
            ids = df.columns            
            uniprots = [list(df[i][df[i].notnull()]) for i in ids]
            uniprots_num = [len(u) for u in uniprots]
            values = [file.split('\\')[4]]*len(ids)
            
            df_outcome = pd.DataFrame({'ID':ids,'Uniprots':uniprots,'Uniprots_num':uniprots_num,'Value':values})                
            df_outcome.to_csv(file,index =False)
           
            print('{} finished'.format(file))
                  

def DB_dic_smiles(path):
# get bingdingDB's cid_smiles dic   
    dic ={}
    files = os.listdir(path) 
    out_files = iter_files(out_path)
    for file in files:
        df = pd.read_csv(os.path.join(path,file))
        dic.update({str(k):v for (k,v) in zip(df['MOL_ID'],df['SMILE'])})
        print('{} dic finished'.format(file))    


def PC_dic_smiles(path_1):
# get Pubchem's cid_smiles dic(after downloading by Webservice)
    files = os.listdir(path_1)
    dic = {}
    for file in files:
        df = pd.read_csv(os.path.join(path_1,file),sep = '\t')
        dic.update({k:v for (k,v) in zip(df.iloc[:,0],df.iloc[:,1])}) 
        print('{} dic finished'.format(file))             


def CB_dic_smiles(path_1,path_2):
# get Chembl's cid_smiles dic,Add "file_fail" because of the Complexity of chembl's origin data
    files_1 = os.listdir(path_1)
    files_2 = os.listdir(path_2)
    out_files = iter_files(out_path)
    dic = {}
    file_fail =[]
    for file_1 in files_1:
        try:
            df = pd.read_csv(os.path.join(path_1,file_1))
            dic.update({str(k):v for (k,v) in zip(df['molecule_chembl_id'],df['canonical_smiles'])})
            print('{} dic finished'.format(file_1))
        except Exception as e:
            print(e)
            file_fail.append(file_1)
            pass
    for file_2 in files_2:
        try:
            df = pd.read_csv(os.path.join(path_2,file_2))
            dic.update({str(k):v for (k,v) in zip(df['Molecule ChEMBL ID'],df['Smiles'])})
            print('{} dic finished'.format(file_2))
        except Exception as e:
            print(e)
            file_fail.append(file_2)
            pass
        
def smile_columns(path)
# Add smile columns for data outcome
	files = iter_files(path)
    for file in files:
       if re.findall('\d+_\d+.csv',file):
           df_out = pd.read_csv(file,sep =',')
           df_out['Smiles'] = df_out['ID'].map(lambda x :dic[str(x)])
           df_out.to_csv(file)
           print('{} finished'.format(file))

因为下载数据里没又inchikey smile不唯一之后的去重会产生问题,所以利用Knime软件计算了inchikey,workflow 如下:

初始的批量文件路径文件记得要在前加上"file:文件路径" 才可以作为变量循环读取.
以文本文件作为file reader的输入,每一行代表一个文件路径.