from __future__ import division,print_function import time,re,sys,os,ast,collections import pandas as pd import pymongo from pymongo import MongoClient from bson.son import SON from bson.codec_options import CodecOptions import re,csv from add_sugar_protein_site_class import add_sugar_protein_site_class from os import listdir def Is_protein(protein_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): #if arg_head overlaps with a protein entity from database, return True for ei in protein_list: if (int(ei[1])>=arg_head_offset[0] and int(ei[1])<=arg_head_offset[1]) \ or (arg_head_offset[0]>=ei[1] and arg_head_offset[0]<=ei[2]): return (True,ei[3],ei[0],ei[4]) #if there is no match, return False return (False,None,'','') def get_protein(protein_list,arg_head_offset): #if arg_head overlaps with a protein entity from database, return offset for ei in protein_list: if (int(ei[1])>=arg_head_offset[0] and int(ei[1])<=arg_head_offset[1]) \ or (arg_head_offset[0]>=ei[1] and arg_head_offset[0]<=ei[2]): return ei[0] def Is_domain(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): domain_name=['sequon','face','head','tail','fragment','lobe','peptide','faceside','domain','motif','region','terminus','epitope','sequence','arm',\ 'sequons','faces','heads','tails','fragments','lobes','peptides','facesides','domains','motifs','regions','termini','epitopes','sequences','arms'] if arg_head.lower() in domain_name: return True else: return False def Is_site(site_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): #if arg_head overlaps with a site entity from database, return True for ei in site_list: if (int(ei[1])>=arg_head_offset[0] and int(ei[1])<=arg_head_offset[1]) \ or (arg_head_offset[0]>=ei[1] and arg_head_offset[0]<=ei[2]): return (True,ei[3],ei[0]) return (False,'','') def get_site(site_list,arg_head_offset): #if arg_head overlaps with a site entity from database, return offset for ei in site_list: if (int(ei[1])>=arg_head_offset[0] and int(ei[1])<=arg_head_offset[1]) \ or (arg_head_offset[0]>=ei[1] and arg_head_offset[0]<=ei[2]): return ei[0] def Is_sugar(sugar_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): #if arg_head overlaps with a sugar entity from database, return True for ei in sugar_list: if (int(ei[1])>=arg_head_offset[0] and int(ei[1])<=arg_head_offset[1]) \ or (arg_head_offset[0]>=ei[1] and arg_head_offset[0]<=ei[2]): return True return False def get_sugar(sugar_list,arg_head_offset): #if arg_head overlaps with a sugar entity from database, return offset for ei in sugar_list: if (int(ei[1])>=arg_head_offset[0] and int(ei[1])<=arg_head_offset[1]) \ or (arg_head_offset[0]>=ei[1] and arg_head_offset[0]<=ei[2]): return ei[0] def generate_new_row(new_row, old_row, old_row_contenct): new_row_content=[] for ni in new_row: if not old_row_contenct[old_row.index(ni)]: new_row_content.append(old_row_contenct[old_row.index(ni)]) else: new_row_content.append(old_row_contenct[old_row.index(ni)].encode('utf-8'))# return new_row_content def contain_digit(string): for si in string: if si.isdigit(): return True return False def process_edg_output2(tsv_input,tsv_output,tsv_output2,tsv_output_c_term,entity_obj,pmidList): col_new = ["doc_id", "sent_index", "relation_id", "relation", "trigger", "trigger_offset", \ "arg_num", "arg_base_np","arg_protein","arg_domain","arg_site","arg_sugar",'PSource','SiteSource','NProtein','NID', 'SiteName',"sent_text"] protein_name_pattern='(.*)glycosylations?$' glyc_pattern='(glyc)' com_glyc_pattern=re.compile(glyc_pattern) with open(tsv_output_c_term, 'w') as csvfile: spamwriter_cterm = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) #write column names spamwriter_cterm.writerow(col_new) with open(tsv_output2, 'w') as csvfile: spamwriter_o2 = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) #write column names spamwriter_o2.writerow(col_new) with open(tsv_output, 'w') as csvfile: spamwriter_o = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) #write column names spamwriter_o.writerow(col_new) with open(tsv_input) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row #print(col_title) break #print(col_title) pmid_index=col_title.index('doc_id') arg_trigger_index=col_title.index('trigger') arg_trigger_offset_index=col_title.index('trigger_offset') arg_head_index=col_title.index('arg_head') arg_head_offset_index=col_title.index('arg_head_offset') arg_base_np_index=col_title.index('arg_base_np') arg_protein_index=col_title.index('arg_protein') arg_domain_index=col_title.index('arg_domain') arg_sugar_index=col_title.index('arg_sugar') arg_site_index=col_title.index('arg_site') arg_base_np_offset_index=col_title.index('arg_base_np_offset') #for relation and arg_n relation_index=col_title.index('relation') arg_num_index=col_title.index('arg_num') sent_index_index=col_title.index('sent_index') relation_id_index=col_title.index('relation_id') #this is for protein source,normallization Source_index=col_title.index('PSource') SiteSource_index=col_title.index('SiteSource') NProtein_index=col_title.index('NProtein') NID_index=col_title.index('NID') SiteName_index=col_title.index('SiteName') #relation dictionary rela_dic_list=collections.defaultdict(list) rela_dic_other={} rela_id_set_contain=set() for row in spamreader: row=[ri.decode('utf-8') for ri in row] rela_dic={} pmid=row[pmid_index] if str(pmid) not in pmidList: continue trigger=row[arg_trigger_index] trigger_offset=row[arg_trigger_offset_index] arg_head=row[arg_head_index] arg_base_np=row[arg_base_np_index] arg_head_offset=row[arg_head_offset_index] arg_base_np_offset=row[arg_base_np_offset_index] #parse the offset to get the real offset tuple trigger_offset_split=trigger_offset.split(':') trigger_offset=(int(trigger_offset_split[0]),int(trigger_offset_split[1])) arg_head_offset_split=arg_head_offset.split(':') arg_head_offset=(int(arg_head_offset_split[0]),int(arg_head_offset_split[1])) arg_base_np_offset_split=arg_base_np_offset.split(':') arg_base_np_offset=(int(arg_base_np_offset_split[0]),int(arg_base_np_offset_split[1])) protein_list,site_list,sugar_list=entity_obj.get_entity_list_from_mongoDB_after(pmid) m_ofin=re.finditer(protein_name_pattern,trigger) count_ofin=[1 for i in m_ofin] count_from=count_ofin #deal with different relations arg1_sign=False arg2_sign=False arg3_sign=False relation=row[relation_index].strip() arg_num=row[arg_num_index].strip() rela_id=str(pmid)+'-'+str(row[sent_index_index])+'-' \ +str(row[relation_id_index])+'-'+relation+'-' #case 1: gly if relation =='gly': is_protein_gly=Is_protein(protein_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset) is_site_gly=Is_site(site_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset) if arg_num=='arg1': if is_protein_gly[0]: row[arg_protein_index]=row[arg_base_np_index] row[Source_index]=is_protein_gly[1] row[NProtein_index]=is_protein_gly[2] row[NID_index]=is_protein_gly[3] row[SiteSource_index]=is_site_gly[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_gly[1]!='Fterm': spamwriter_cterm.writerow(new_row) elif is_site_gly[0] or \ Is_domain(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): row[arg_site_index]=row[arg_base_np_index] row[SiteName_index]=is_site_gly[2] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) elif Is_sugar(sugar_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): row[arg_sugar_index]=row[arg_base_np_index] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) elif arg_num=='arg2': if is_site_gly[0] or \ Is_domain(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): row[arg_site_index]=row[arg_base_np_index] row[SiteName_index]=is_site_gly[2] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) elif arg_num=='arg3': if Is_sugar(sugar_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset): row[arg_sugar_index]=row[arg_base_np_index] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) #case 2: attach elif relation =='attach': rela_dic_list[rela_id+arg_num].append((arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset)) if rela_id+'arg1' in rela_dic_list and rela_id+'arg2' in rela_dic_list and arg_num in ['arg1','arg2']: if arg_num=='arg1': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg1_sign=True elif arg_num=='arg2': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg2_sign=True arg_1_str=rela_id+'arg1' arg_2_str=rela_id+'arg2' if arg1_sign: list_arg=rela_dic_list[arg_2_str] elif arg2_sign: list_arg=rela_dic_list[arg_1_str] for li in list_arg: if arg1_sign: rela_dic[arg_2_str]=li elif arg2_sign: rela_dic[arg_1_str]=li is_protein_attach=Is_protein(protein_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_attach_arg1=Is_site(site_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) if ( is_site_attach_arg1[0] or \ Is_domain(rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]))\ and Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=rela_dic[arg_1_str][2] row[SiteName_index]=is_site_attach_arg1[2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('attach:',new_row) elif is_protein_attach[0] and \ Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' row[Source_index]=is_protein_attach[1] row[NProtein_index]=is_protein_attach[2] row[NID_index]=is_protein_attach[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_attach[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('attach:',new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) if rela_id+'arg2' in rela_dic_list and rela_id+'arg3' in rela_dic_list and arg_num in ['arg2','arg3']: if arg_num=='arg2': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg2_sign=True elif arg_num=='arg3': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg3_sign=True arg_1_str=rela_id+'arg2' arg_2_str=rela_id+'arg3' if arg2_sign: list_arg=rela_dic_list[arg_2_str] elif arg3_sign: list_arg=rela_dic_list[arg_1_str] for li in list_arg: if arg2_sign: rela_dic[arg_2_str]=li elif arg3_sign: rela_dic[arg_1_str]=li is_site_attach_arg2=Is_site(site_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]) if (is_site_attach_arg2[0] or \ Is_domain(rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3])) \ and Is_sugar(sugar_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=rela_dic[arg_2_str][2] row[SiteName_index]=is_site_attach_arg2[2] row[arg_sugar_index]=rela_dic[arg_1_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('attach:',new_row) #case 2: contain elif relation =='contain': rela_dic_list[rela_id+arg_num].append((arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset,tuple(row))) rela_id_set_contain.add(rela_id) #case 3: located elif relation =='located': rela_dic_list[rela_id+arg_num].append((arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset)) if rela_id+'arg1' in rela_dic_list and rela_id+'arg2' in rela_dic_list: if arg_num=='arg1': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg1_sign=True elif arg_num=='arg2': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg2_sign=True arg_1_str=rela_id+'arg1' arg_2_str=rela_id+'arg2' if arg1_sign: list_arg=rela_dic_list[arg_2_str] elif arg2_sign: list_arg=rela_dic_list[arg_1_str] for li in list_arg: if arg1_sign: rela_dic[arg_2_str]=li elif arg2_sign: rela_dic[arg_1_str]=li is_protein_located=Is_protein(protein_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_located_arg1=Is_site(site_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_located_arg2=Is_site(site_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]) if (is_site_located_arg1[0] or \ Is_domain(rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3])) \ and Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=rela_dic[arg_1_str][2] row[SiteName_index]=is_site_located_arg1[2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('located:',new_row) elif is_protein_located[0] and \ Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' row[Source_index]=is_protein_located[1] row[NProtein_index]=is_protein_located[2] row[NID_index]=is_protein_located[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_located[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('located:',new_row) elif (is_site_located_arg2[0] or \ Is_domain(rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3])) and \ is_protein_located[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_site_index]=rela_dic[arg_2_str][2] row[SiteName_index]=is_site_located_arg2[2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='part_of' row[Source_index]=is_protein_located[1] row[NProtein_index]=is_protein_located[2] row[NID_index]=is_protein_located[3] row[SiteSource_index]=is_site_located_arg2[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_located[1]!='Fterm' and is_site_located_arg2[0] \ and (is_site_located_arg2[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_located_arg2[2]): spamwriter_cterm.writerow(new_row) #print('located:',new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) #case 4: theme elif relation =='theme': rela_dic_other[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg_1_str=rela_id+'arg1' is_protein_theme=Is_protein(protein_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_theme_trigger=Is_site(site_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_theme_arg1=Is_site(site_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) if (is_site_theme_trigger[0] or \ Is_domain(rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3])) and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=get_site(site_list,rela_dic_other[arg_1_str][5]) row[SiteName_index]=is_site_theme_trigger[2] row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('theme:',new_row) elif is_protein_theme[0] and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_theme[1] row[NProtein_index]=is_protein_theme[2] row[NID_index]=is_protein_theme[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_theme[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('theme:',new_row) elif (is_site_theme_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) ) and \ is_protein_theme[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_site_index]=rela_dic_other[arg_1_str][2] row[SiteName_index]=is_site_theme_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='part_of' row[Source_index]=is_protein_theme[1] row[NProtein_index]=is_protein_theme[2] row[NID_index]=is_protein_theme[3] row[SiteSource_index]=is_site_theme_arg1[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_theme[1]!='Fterm' and is_site_theme_arg1[0] \ and (is_site_theme_arg1[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_theme_arg1[2]): spamwriter_cterm.writerow(new_row) #print('theme:',new_row) #case 5: from elif relation =='from': rela_dic_other[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg_1_str=rela_id+'arg1' is_protein_from_arg1=Is_protein(protein_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_protein_from_trigger=Is_protein(protein_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_from_trigger=Is_site(site_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_from_arg1=Is_site(site_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) if (is_site_from_trigger[0] or \ Is_domain(rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3])) and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=get_site(site_list,rela_dic_other[arg_1_str][5]) row[SiteName_index]=is_site_from_trigger[2] row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('from:',new_row) elif is_protein_from_trigger[0] and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_from_trigger[1] row[NProtein_index]=is_protein_from_trigger[2] row[NID_index]=is_protein_from_trigger[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_from_trigger[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('from:',new_row) elif Is_sugar(sugar_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) and \ is_protein_from_arg1[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][1]) row[arg_sugar_index]=rela_dic_other[arg_1_str][4] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_from_arg1[1] row[NProtein_index]=is_protein_from_arg1[2] row[NID_index]=is_protein_from_arg1[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_from_arg1[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('from:',new_row) elif (is_site_from_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) ) and \ is_protein_from_trigger[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_site_index]=rela_dic_other[arg_1_str][2] row[SiteName_index]=is_site_from_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='part_of' row[Source_index]=is_protein_from_trigger[1] row[NProtein_index]=is_protein_from_trigger[2] row[NID_index]=is_protein_from_trigger[3] row[SiteSource_index]=is_site_from_arg1[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_from_trigger[1]!='Fterm' and is_site_from_arg1[0] \ and (is_site_from_arg1[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_from_arg1[2]): spamwriter_cterm.writerow(new_row) #print('from:',new_row) ###Need to work on this !!!!!! elif (is_site_from_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) ) and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=rela_dic_other[arg_1_str][2] row[SiteName_index]=is_site_from_arg1[2] row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[SiteSource_index]=is_site_from_arg1[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_from_trigger[1]!='Fterm' and is_site_from_arg1[0] \ and (is_site_from_arg1[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_from_arg1[2]): spamwriter_cterm.writerow(new_row) #print('from:',new_row) elif (is_site_from_trigger[0] or \ Is_domain(rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) ) and \ is_protein_from_arg1[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][1]) row[arg_site_index]=rela_dic_other[arg_1_str][4] row[SiteName_index]=is_site_from_trigger[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='part_of' row[Source_index]=is_protein_from_arg1[1] row[NProtein_index]=is_protein_from_arg1[2] row[NID_index]=is_protein_from_arg1[3] row[SiteSource_index]=is_site_from_trigger[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_from_arg1[1]!='Fterm' and is_site_from_trigger[0] \ and (is_site_from_trigger[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_from_trigger[2]): spamwriter_cterm.writerow(new_row) #print('from:',new_row) elif is_site_from_arg1[0] and \ Is_domain(rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) : #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic_other[arg_1_str][4] row[arg_site_index]=rela_dic_other[arg_1_str][2] row[SiteName_index]=is_site_from_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='part_of' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('from:',new_row) elif (is_site_from_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3])) and \ sum(count_from)>0: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=get_site(site_list,rela_dic_other[arg_1_str][1]) row[SiteName_index]=is_site_from_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('from:',new_row) #case 6: with elif relation =='with': rela_dic_other[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg_1_str=rela_id+'arg1' is_protein_with=Is_protein(protein_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_with_trigger=Is_site(site_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_with_arg1=Is_site(site_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) if (is_site_with_trigger[0] or \ Is_domain(rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3])) and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=get_site(site_list,rela_dic_other[arg_1_str][5]) row[SiteName_index]=is_site_with_trigger[2] row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('with:',new_row) elif is_protein_with[0] and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_with[1] row[NProtein_index]=is_protein_with[2] row[NID_index]=is_protein_with[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_with[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('with:',new_row) elif (is_site_with_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) ) and \ is_protein_with[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_site_index]=rela_dic_other[arg_1_str][2] row[SiteName_index]=is_site_with_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='part_of' row[Source_index]=is_protein_with[1] row[NProtein_index]=is_protein_with[2] row[NID_index]=is_protein_with[3] row[SiteSource_index]=is_site_with_arg1[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_with[1]!='Fterm' and is_site_with_arg1[0] \ and (is_site_with_arg1[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_with_arg1[2]): spamwriter_cterm.writerow(new_row) #print('with:',new_row) elif is_site_with_arg1[0] and \ Is_domain(rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) : #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic_other[arg_1_str][4] row[arg_site_index]=rela_dic_other[arg_1_str][2] row[SiteName_index]=is_site_with_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='part_of' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('with:',new_row) #case 6: link elif relation =='link': rela_dic_other[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg_1_str=rela_id+'arg1' is_protein_link=Is_protein(protein_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) if is_protein_link[0] and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][5]) row[arg_sugar_index]=rela_dic_other[arg_1_str][2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_link[1] row[NProtein_index]=is_protein_link[2] row[NID_index]=is_protein_link[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_link[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('link:',new_row) #case 7: mod elif relation =='mod': rela_dic_list[rela_id+arg_num].append((arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset)) if rela_id+'arg1' in rela_dic_list and rela_id+'arg3' in rela_dic_list and arg_num in ['arg1','arg3']: if arg_num=='arg1': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg1_sign=True elif arg_num=='arg3': rela_dic[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg2_sign=True arg_1_str=rela_id+'arg1' arg_2_str=rela_id+'arg3' if arg1_sign: list_arg=rela_dic_list[arg_2_str] elif arg2_sign: list_arg=rela_dic_list[arg_1_str] for li in list_arg: if arg1_sign: rela_dic[arg_2_str]=li elif arg2_sign: rela_dic[arg_1_str]=li is_protein_mod=Is_protein(protein_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_mod_arg1=Is_site(site_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) if (is_site_mod_arg1[0] or \ Is_domain(rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3])) \ and Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=rela_dic[arg_1_str][2] row[SiteName_index]=is_site_mod_arg1[2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('mod:',new_row) elif is_protein_mod[0] and \ Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' row[Source_index]=is_protein_mod[1] row[NProtein_index]=is_protein_mod[2] row[NID_index]=is_protein_mod[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_mod[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('mod:',new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) #case 8: ofin elif relation =='ofin': rela_dic_other[rela_id+arg_num]=(arg_head,arg_head_offset,arg_base_np,arg_base_np_offset,trigger,trigger_offset) arg_1_str=rela_id+'arg1' is_protein_ofin=Is_protein(protein_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) is_site_ofin_arg1=Is_site(site_list,rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]) if (is_site_ofin_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3])) and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=get_site(site_list,rela_dic_other[arg_1_str][1]) row[SiteName_index]=is_site_ofin_arg1[2] row[arg_sugar_index]=rela_dic_other[arg_1_str][4] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('ofin:',new_row) elif is_protein_ofin[0] and \ Is_sugar(sugar_list,rela_dic_other[arg_1_str][4],rela_dic_other[arg_1_str][5],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][1]) row[arg_sugar_index]=rela_dic_other[arg_1_str][4] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_ofin[1] row[NProtein_index]=is_protein_ofin[2] row[NID_index]=is_protein_ofin[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_ofin[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('ofin:',new_row) elif (is_site_ofin_arg1[0] or \ Is_domain(rela_dic_other[arg_1_str][0],rela_dic_other[arg_1_str][1],rela_dic_other[arg_1_str][2],rela_dic_other[arg_1_str][3])) and \ sum(count_ofin)>0: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=get_site(site_list,rela_dic_other[arg_1_str][1]) row[SiteName_index]=is_site_ofin_arg1[2] row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('ofin:',new_row) elif is_protein_ofin[0] and sum(count_ofin)>0: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=get_protein(protein_list,rela_dic_other[arg_1_str][1]) row[arg_base_np_index]=rela_dic_other[arg_1_str][2] row[relation_index]='gly' row[Source_index]=is_protein_ofin[1] row[NProtein_index]=is_protein_ofin[2] row[NID_index]=is_protein_ofin[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_ofin[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('ofin:',new_row) elif relation =='dep': is_site_dep=Is_site(site_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset) raw_doc=entity_obj.get_abstract_text(pmid) #match the glyc text if raw_doc: abstract_text=raw_doc["text"] sentence = raw_doc["sentence"] for senInfo in sentence: senIndex = senInfo["index"] if str(row[sent_index_index])==str(senIndex): sent_start_of_site=int(senInfo["charStart"]) break text_before_site=abstract_text[sent_start_of_site:arg_head_offset[0]+1] sr=com_glyc_pattern.search(text_before_site) if is_site_dep[0] and sr: row[arg_site_index]=row[arg_base_np_index] row[SiteName_index]=is_site_dep[2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) elif relation =='used': is_site_used=Is_site(site_list,arg_head,arg_head_offset,arg_base_np,arg_base_np_offset) if is_site_used[0]: row[arg_site_index]=row[arg_base_np_index] row[SiteName_index]=is_site_used[2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #for the contain case for rela_id in list(rela_id_set_contain): pmid=rela_id.split('-')[0] protein_list,site_list,sugar_list=entity_obj.get_entity_list_from_mongoDB_after(pmid) row=rela_dic_list arg_1_str=rela_id+'arg1' arg_2_str=rela_id+'arg2' arg_3_str=rela_id+'arg3' if arg_1_str in rela_dic_list: row=rela_dic_list[arg_1_str][0][6] elif arg_2_str in rela_dic_list: row=rela_dic_list[arg_2_str][0][6] elif arg_3_str in rela_dic_list: row=rela_dic_list[arg_3_str][0][6] row=list(row) rela_dic={} if rela_id+'arg1' in rela_dic_list and rela_id+'arg2' in rela_dic_list and rela_id+'arg3' in rela_dic_list: #print('contain3:',rela_id,',',rela_dic_list[arg_1_str],',',rela_dic_list[arg_2_str],',',rela_dic_list[arg_3_str]) for arg1_i in rela_dic_list[arg_1_str]: for arg2_i in rela_dic_list[arg_2_str]: for arg3_i in rela_dic_list[arg_3_str]: rela_dic[arg_1_str]=arg1_i rela_dic[arg_2_str]=arg2_i rela_dic[arg_3_str]=arg3_i is_protein_contain=Is_protein(protein_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_contain_arg3=Is_site(site_list,rela_dic[arg_3_str][0],rela_dic[arg_3_str][1],rela_dic[arg_3_str][2],rela_dic[arg_3_str][3]) if is_protein_contain[0] and \ Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]) and \ is_site_contain_arg3[0]: #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_site_index]=rela_dic[arg_3_str][2] row[SiteName_index]=is_site_contain_arg3[2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' row[Source_index]=is_protein_contain[1] row[NProtein_index]=is_protein_contain[2] row[NID_index]=is_protein_contain[3] row[SiteSource_index]=is_site_contain_arg3[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_contain[1]!='Fterm' and is_site_contain_arg3[0] \ and (is_site_contain_arg3[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_contain_arg3[2]): spamwriter_cterm.writerow(new_row) #print('contain3:',new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) elif rela_id+'arg1' in rela_dic_list and rela_id+'arg2' in rela_dic_list: for arg1_i in rela_dic_list[arg_1_str]: for arg2_i in rela_dic_list[arg_2_str]: rela_dic[arg_1_str]=arg1_i rela_dic[arg_2_str]=arg2_i is_protein_contain=Is_protein(protein_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_contain_arg1=Is_site(site_list,rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]) is_site_contain_arg2=Is_site(site_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]) if (is_site_contain_arg1[0] or \ Is_domain(rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3])) and \ Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_site_index]=rela_dic[arg_1_str][2] row[SiteName_index]=is_site_contain_arg1[2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('contain2:',new_row) elif is_protein_contain[0] and \ Is_sugar(sugar_list,rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_sugar_index]=rela_dic[arg_2_str][2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='gly' row[Source_index]=is_protein_contain[1] row[NProtein_index]=is_protein_contain[2] row[NID_index]=is_protein_contain[3] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_contain[1]!='Fterm': spamwriter_cterm.writerow(new_row) #print('contain2:',new_row) elif is_protein_contain[0] and \ (is_site_contain_arg2[0] or \ Is_domain(rela_dic[arg_2_str][0],rela_dic[arg_2_str][1],rela_dic[arg_2_str][2],rela_dic[arg_2_str][3])): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_site_index]=rela_dic[arg_2_str][2] row[SiteName_index]=is_site_contain_arg2[2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='part_of' row[Source_index]=is_protein_contain[1] row[NProtein_index]=is_protein_contain[2] row[NID_index]=is_protein_contain[3] row[SiteSource_index]=is_site_contain_arg2[1] new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) if is_protein_contain[1]!='Fterm' and is_site_contain_arg2[0] \ and (is_site_contain_arg2[1]in ['AminoAcid','SpecificSite','SiteSequence']) and \ contain_digit(is_site_contain_arg2[2]): spamwriter_cterm.writerow(new_row) #print('contain2:',new_row) elif is_site_contain_arg2[0] and \ Is_domain(rela_dic[arg_1_str][0],rela_dic[arg_1_str][1],rela_dic[arg_1_str][2],rela_dic[arg_1_str][3]): #how to update the row content: arg_num, arg_head, arg_base_np, etc? row[arg_protein_index]=rela_dic[arg_1_str][2] row[arg_site_index]=rela_dic[arg_2_str][2] row[SiteName_index]=is_site_contain_arg2[2] row[arg_base_np_index]=rela_dic[arg_1_str][2]+' AND '+rela_dic[arg_2_str][2] row[relation_index]='part_of' new_row=generate_new_row(col_new,col_title,row) spamwriter_o.writerow(new_row) #print('contain2:',new_row) else: new_row=generate_new_row(col_new,col_title,row) spamwriter_o2.writerow(new_row) def propagation_rules(input_file,output_file): with open(output_file, 'w') as csvfile: spamwriter_o = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) part_of_dic={} with open(input_file) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row #print(col_title) break #write column names spamwriter_o.writerow(col_title) #print(col_title) pmid_index=col_title.index('doc_id') arg_protein_index=col_title.index('arg_protein') arg_site_index=col_title.index('arg_site') #for relation and arg_n relation_index=col_title.index('relation') arg_num_index=col_title.index('arg_num') sent_index_index=col_title.index('sent_index') relation_id_index=col_title.index('relation_id') for row in spamreader: pmid=row[pmid_index] #deal with different relations relation=row[relation_index].strip() rela_id=str(pmid)+'-'+str(row[sent_index_index])+'-' \ +relation+'-' if relation == 'part_of': part_of_dic[rela_id]=(row[arg_site_index],row[arg_protein_index]) #print(part_of_dic) with open(input_file) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row break #print(col_title) pmid_index=col_title.index('doc_id') arg_protein_index=col_title.index('arg_protein') arg_domain_index=col_title.index('arg_domain') arg_sugar_index=col_title.index('arg_sugar') arg_site_index=col_title.index('arg_site') #for relation and arg_n relation_index=col_title.index('relation') arg_num_index=col_title.index('arg_num') sent_index_index=col_title.index('sent_index') relation_id_index=col_title.index('relation_id') for row in spamreader: pmid=row[pmid_index] #deal with different relations relation=row[relation_index].strip() arg_num=row[arg_num_index].strip() rela_id=str(pmid)+'-'+str(row[sent_index_index])+'-' \ +'part_of-' if relation == 'gly' and rela_id in part_of_dic and row[arg_site_index]==part_of_dic[rela_id][0]: row[arg_protein_index]=part_of_dic[rela_id][1] #print(row) spamwriter_o.writerow(row) def merge_rows(input_file,output_file): with open(output_file, 'w') as csvfile: spamwriter_o = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) row_dic={} merged_row_dic={} with open(input_file) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row #print(col_title) break #write column names spamwriter_o.writerow(col_title) #print(col_title) pmid_index=col_title.index('doc_id') arg_protein_index=col_title.index('arg_protein') arg_site_index=col_title.index('arg_site') arg_sugar_index=col_title.index('arg_sugar') #for relation and arg_n relation_index=col_title.index('relation') arg_num_index=col_title.index('arg_num') sent_index_index=col_title.index('sent_index') trigger_offset_index=col_title.index('trigger_offset') #this is for protein source,normallization Source_index=col_title.index('PSource') SiteSource_index=col_title.index('SiteSource') NProtein_index=col_title.index('NProtein') NID_index=col_title.index('NID') SiteName_index=col_title.index('SiteName') for row in spamreader: row=[ri.decode('utf-8') for ri in row ] pmid=row[pmid_index] #for test purpose and remove this after test #row[-1]='' #print('PMID:',pmid) #deal with different relations relation=row[relation_index].strip() rela_id=str(pmid)+'-'+str(row[sent_index_index])+'-' \ +relation+'-' if rela_id in row_dic: row_dic[rela_id].append(row) else: row_dic[rela_id]=[row] #output the data for ri in row_dic: for rii in range(len(row_dic[ri])): cur_row=tuple(row_dic[ri][rii]) #check if they need to merge for riii in range(rii,len(row_dic[ri])): if riii==rii: continue loop_row=tuple(row_dic[ri][riii]) if loop_row[trigger_offset_index]==cur_row[trigger_offset_index]: protein_cmp=compatible(loop_row[NProtein_index],cur_row[NProtein_index]) site_cmp=compatible(loop_row[SiteName_index],cur_row[SiteName_index]) sugar_cmp=compatible(loop_row[arg_sugar_index],cur_row[arg_sugar_index]) #print('Current row:',row[:-1]) #print('Rows in dic:',row_dic[rela_id]) #print('Protein_cmp:',protein_cmp,'--','Site_cmp:',site_cmp,'--','sugar_cmp:',sugar_cmp) if protein_cmp[0] and site_cmp[0] and sugar_cmp[0]: merged_row=list(loop_row) merged_row[arg_protein_index]=protein_cmp[1] merged_row[arg_site_index]=site_cmp[1] merged_row[arg_sugar_index]=sugar_cmp[1] #update the important information merged_row[Source_index]=compatible(merged_row[Source_index],cur_row[Source_index])[1] merged_row[SiteSource_index]=compatible(merged_row[SiteSource_index],cur_row[SiteSource_index])[1] merged_row[NProtein_index]=compatible(merged_row[NProtein_index],cur_row[NProtein_index])[1] merged_row[NID_index]=compatible(merged_row[NID_index],cur_row[NID_index])[1] merged_row[SiteName_index]=compatible(merged_row[SiteName_index],cur_row[SiteName_index])[1] if ri in merged_row_dic: merged_row_dic[ri].append(merged_row) else: merged_row_dic[ri]=[merged_row] for ri in row_dic: for rii in row_dic[ri]: already_merged=False #skip the merged rows: if ri in merged_row_dic: for rmi in merged_row_dic[ri]: if rmi[trigger_offset_index]==rii[trigger_offset_index]: protein_cmp=compatible(rmi[NProtein_index],rii[NProtein_index]) site_cmp=compatible(rmi[SiteName_index],rii[SiteName_index]) sugar_cmp=compatible(rmi[arg_sugar_index],rii[arg_sugar_index]) #print('Current row:',row[:-1]) #print('Rows in dic:',row_dic[rela_id]) #print('Protein_cmp:',protein_cmp,'--','Site_cmp:',site_cmp,'--','sugar_cmp:',sugar_cmp) if protein_cmp[0] and site_cmp[0] and sugar_cmp[0]: already_merged=True break if already_merged: continue rii=[rri.encode('utf-8') for rri in rii] spamwriter_o.writerow(rii) if ri in merged_row_dic: for rii in merged_row_dic[ri]: rii=[rri.encode('utf-8') for rri in rii] spamwriter_o.writerow(rii) def compatible(e1,e2): if e1 and e2 and e1==e2: return (True, e1) elif e1 and e2 and e1!=e2: return (False,'') elif e1 and not e2: return (True,e1) elif e2 and not e1: return (True,e2) elif not e1 and not e2: return (True,'') def extrac_certain_columns_from_postprocessed_file(input_file,output_file): col_new = ["doc_id", "sent_index","trigger","PSource", "NProtein",'NID','SiteName'] with open(output_file, 'w') as csvfile: spamwriter_o = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) #write column names spamwriter_o.writerow(col_new) with open(input_file) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row relation_index=col_title.index('relation') break for row in spamreader: row=[ri.decode('utf-8') for ri in row ] if row[relation_index]=='part_of': continue #row=[str(ri, 'utf-8', 'ignore') for ri in row ] new_row=generate_new_row(col_new, col_title, row) #new_row=[ri.encode('utf-8') for ri in new_row ] spamwriter_o.writerow(new_row) def add_columns(file_input,added_file,columns_list): with open(added_file, 'w') as csvfile: spamwriter_o = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) with open(file_input) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row #print(col_title) break #write column names col_title=col_title+columns_list spamwriter_o.writerow(col_title) for row in spamreader: #row=[ri.decode('utf-8') for ri in row ] row=row+['']*len(columns_list) spamwriter_o.writerow(row) def substract_file(file1,file2,file3): with open(file3, 'w') as csvfile: spamwriter_o = csv.writer(csvfile, delimiter='\t', quotechar='|', quoting=csv.QUOTE_MINIMAL) row_list1=[] row_list2=[] with open(file1) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row #print(col_title) break for row in spamreader: #row=[ri.decode('utf-8') for ri in row ] row_list1.append(row) with open(file2) as csvfile: spamreader = csv.reader(csvfile, delimiter='\t', quotechar='|') for row in spamreader: col_title=row #print(col_title) break for row in spamreader: #row=[ri.decode('utf-8') for ri in row ] row_list2.append(row) #write column names spamwriter_o.writerow(col_title) for ri in row_list1: if ri not in row_list2: #ri=[rri.encode('utf-8') for rri in ri ] spamwriter_o.writerow(ri) if __name__ == '__main__': #test pmidFile = sys.argv[1] dbF = sys.argv[2] dbT = sys.argv[3] colF = sys.argv[4] colT = sys.argv[5] textT = sys.argv[6] edg = sys.argv[7] ASPSC=add_sugar_protein_site_class(dbF,dbT,colF,colT,textT,textT) tsv_input=edg pmidList = pd.read_csv(pmidFile,header=None).iloc[:,0].tolist() pmidList=[str(i) for i in pmidList] tsv_input_split=tsv_input.split('.') tsv_output=tsv_input_split[0]+'_postprocessing_before_merge.tsv' tsv_output2=tsv_input_split[0]+'_filtered_out_before_merge.tsv' tsv_output_cterm=tsv_input_split[0]+'_cterm_before_merge.tsv' tsv_centain_colums=tsv_input_split[0]+'_output.tsv' tsv_input_new=tsv_input_split[0]+'_new.tsv' #add some columns on the files add_columns(tsv_input,tsv_input_new,['PSource','NProtein','NID','SiteSource','SiteName']) process_edg_output2(tsv_input_new,tsv_output,tsv_output2,tsv_output_cterm,ASPSC,pmidList) #extrac_certain_columns_from_postprocessed_file(tsv_output,tsv_centain_colums) #tsv_output_new='glyco_postprocessing_new.tsv' #propagation_rules(tsv_output,tsv_output_new) output_file=tsv_input_split[0]+'_postprocessing.tsv' output_file2=tsv_input_split[0]+'_filtered_out.tsv' output_file_cterm=tsv_input_split[0]+'_cterm.tsv' output_file_other=tsv_input_split[0]+'_other.tsv' final_file=tsv_input_split[0]+'_final.tsv' merge_rows(tsv_output,output_file) merge_rows(tsv_output2,output_file2) merge_rows(tsv_output_cterm,output_file_cterm) substract_file(output_file,output_file_cterm,output_file_other) extrac_certain_columns_from_postprocessed_file(output_file,final_file) ''' #final step tsv_input_split=['unicarb'] output_file_cterm=tsv_input_split[0]+'_cterm.tsv' output_file_other=tsv_input_split[0]+'_other.tsv' final_file=tsv_input_split[0]+'_final.tsv' extrac_certain_columns_from_postprocessed_file(output_file_cterm,final_file) '''