def analysisBS (name_lig, ID_seq = '0.0', debug = 1): pr_result = pathManage.result(name_lig) pr_out = pathManage.result(name_lig + "/sameBS") # log files p_log_file = pr_out + "log.txt" filout_log = open (p_log_file, "w") # dictionnar with files d_file_BS = {} d_file_BS["global"] = open (pr_out + name_lig + "_", "w") d_file_BS["global"].write ("name_bs\tRMSD_prot\tRMSD_BS_ca\tRMSD_BS_all\tD_max\tl_at_BS\tidentic\n") d_file_BS["summary"] = open (pr_out + "summary.txt", "w") pr_dataset = pathManage.dataset(name_lig) l_folder_ref = listdir(pr_result) nb_BS = 0 nb_BS_filtered = 0 nb_same_BS = 0 for PDB_ref in l_folder_ref : if debug : print PDB_ref if len (PDB_ref) != 4 : continue p_pdb_ref = pathManage.findPDBRef(pr_dataset + PDB_ref + "/") l_p_query = pathManage.findPDBQueryTransloc (pathManage.result(name_lig) + PDB_ref + "/") if debug : print l_p_query for p_query in l_p_query : # read TM Align if debug : print p_query.split ("/")[-1][7:-4] p_TMalign = pathManage.alignmentOutput(name_lig) + p_pdb_ref.split ("/")[-1][0:-4] + "__" + p_query.split ("/")[-1][7:-4] + "/RMSD" try : score_align = parseTMalign.parseOutputTMalign(p_TMalign) except : filout_log.write ("ERROR TM align " + p_TMalign + "\n") continue nb_BS = nb_BS + 1 if score_align["IDseq"] >= ID_seq : nb_BS_filtered = nb_BS_filtered + 1 l_p_substruct_ref = pathManage.findSubstructRef (pr_dataset + PDB_ref + "/", name_lig) # sub BS for p_substruct_ref in l_p_substruct_ref : struct_substitued = p_substruct_ref.split ("_")[-2] # write header if not struct_substitued in d_file_BS.keys () : d_file_BS[struct_substitued] = open (pr_out + name_lig + "_" + struct_substitued + "_", "w") d_file_BS[struct_substitued].write ("name_bs\tRMSD_prot\tRMSD_BS_ca\tRMSD_BS_all\tD_max\tl_at_BS\tidentic\n") RMSD_bs = analysis.computeRMSDBS (p_pdb_ref, p_query, p_substruct_ref, pr_out) if RMSD_bs != [] : d_file_BS[struct_substitued].write (p_substruct_ref.split("/")[-1][0:-4] + "_*_" + p_query.split ("/")[-1][0:-4] + "\t" + str(score_align["RMSD"]) + "\t" + str(RMSD_bs[1]) + "\t" + str(RMSD_bs[0]) + "\t" + str(RMSD_bs[2]) + "\t" + str(RMSD_bs[-2]) + "\t" + str(RMSD_bs[-1]) + "\n") p_ligand_ref = pathManage.findligandRef(pr_dataset + PDB_ref + "/", name_lig) RMSD_bs_lig = analysis.computeRMSDBS (p_pdb_ref, p_query, p_ligand_ref, pr_out) if RMSD_bs_lig != [] : d_file_BS["global"].write (p_ligand_ref.split("/")[-1][0:-4] + "_*_" + p_query.split ("/")[-1][0:-4] + "\t" + str(score_align["RMSD"]) + "\t" + str(RMSD_bs_lig[1]) + "\t" + str(RMSD_bs_lig[0]) + "\t" + str(RMSD_bs_lig[2]) + "\t" + str(RMSD_bs_lig[-2]) + "\t" + str(RMSD_bs_lig[-1]) + "\n") if RMSD_bs_lig [-1] == 1 : nb_same_BS = nb_same_BS + 1 # write summary d_file_BS["summary"].write ("BS global: " + str (nb_BS) + "\n") d_file_BS["summary"].write ("BS - IDseq " + str (ID_seq) + "%: " + str (nb_BS_filtered) + "\n") d_file_BS["summary"].write ("BS - same atom number: " + str (nb_same_BS) + "\n") filout_log.close () # close files and run histograms for k_dico in d_file_BS.keys () : p_file = d_file_BS[k_dico].name d_file_BS[k_dico].close () if name_lig == "ATP" : runOtherSoft.RhistogramRMSD(p_file, max_RMSD = 5.0) elif name_lig == "ADP" : runOtherSoft.RhistogramRMSD(p_file, max_RMSD = 4.0) elif name_lig == "AMP" : runOtherSoft.RhistogramRMSD(p_file, max_RMSD = 4.0) else : runOtherSoft.RhistogramRMSD(p_file, max_RMSD = 3.5) return 1
def retrieveSubstructSuperimposed (name_lig, thresold_BS = 4.5, thresold_superimposed_ribose = 2.5, thresold_superimposed_pi = 3, thresold_shaep = 0.4): # ouput p_dir_dataset = pathManage.dataset(name_lig) p_dir_result = pathManage.result(name_lig ) l_folder_ref = listdir(p_dir_dataset) # log control p_log = open(p_dir_result + "log_superimposed.txt", "w") # control extraction d_control = {} d_control["pr ref"] = 0 d_control["lig query"] = 0 d_control["subref"] = {} d_control["subref empty"] = {} d_control["out sheap"] = {} filout_control = open (p_dir_result + "quality_extraction.txt", "w") # stock smile code d_smile = {} # sheap control d_filout_sheap = {} d_filout_sheap ["list"] = [p_dir_result + "shaep_global.txt"] d_filout_sheap["global"] = open (p_dir_result + "shaep_global.txt", "w") d_filout_sheap["global"].write ("name\tbest_similarity\tshape_similarity\tESP_similarity\n") for ref_folder in l_folder_ref : # control folder reference name if len (ref_folder) != 4 : p_log.write ("[ERROR folder] -> " + ref_folder + "\n") continue # reference p_lig_ref = pathManage.findligandRef(p_dir_dataset + ref_folder + "/", name_lig) try: lig_ref_parsed = parsePDB.loadCoordSectionPDB(p_lig_ref, "HETATM") # print len (lig_ref_parsed) except: p_log.write ("[ERROR ligand ref] -> " + p_lig_ref + "\n") continue #control d_control["pr ref"] = d_control["pr ref"] + 1 # output by reference p_dir_result_ref = pathManage.result(name_lig + "/" + ref_folder) d_filout_superimposed = {} d_filout_superimposed["global"] = open (p_dir_result_ref + "all_ligand_aligned.pdb", "w") d_filout_superimposed["sheap"] = open (p_dir_result_ref + "all_ligand_aligned_" + str (thresold_shaep) + ".pdb", "w") # write lig ref -> connect matrix corrrect in all reference and all sheap writePDBfile.coordinateSection(d_filout_superimposed["global"], lig_ref_parsed, "HETATM", connect_matrix = 1) writePDBfile.coordinateSection(d_filout_superimposed["sheap"], lig_ref_parsed, "HETATM", connect_matrix = 1) # inspect folder dataset l_pdbfile = listdir(p_dir_dataset + ref_folder + "/") for pdbfile in l_pdbfile : # no ligand file if len (pdbfile.split ("_")) == 1 : continue pdbfile = pdbfile[:-4] # remove extention if len(pdbfile.split ("_")[0]) == 3 and len(pdbfile.split ("_")[1]) == 4 and pdbfile.split ("_")[1] != ref_folder: p_lig = p_dir_dataset + ref_folder + "/" + pdbfile + ".pdb" if p_lig_ref != p_lig : # pass case where ligand replace same ligand -> does not need run if pdbfile.split ("_")[0] == name_lig : p_log.write ("[REMOVE] -> same ligand substituate") continue # parsed ligand query lig_parsed = parsePDB.loadCoordSectionPDB(p_lig, "HETATM") # find matrix of rotation p_matrix = pathManage.findMatrix(p_lig_ref, p_lig, name_lig) # control file matrix exist if not path.exists(p_matrix) : p_log.write ("[ERROR] -> Matrix transloc " + p_lig_ref + " " + p_lig + " " + name_lig + "\n") continue # control d_control["lig query"] = d_control["lig query"] + 1 # find the path of complex used p_complex = p_dir_dataset + ref_folder + "/" + p_lig.split ("/")[-1][4:] # ligand rotated -> change the referentiel superposeStructure.applyMatrixLigand(lig_parsed, p_matrix) # use substruct l_p_substruct_ref = pathManage.findSubstructRef (pathManage.dataset(name_lig) + ref_folder + "/" , name_lig) for p_substruct_ref in l_p_substruct_ref : # ribose or phosphate struct_type = p_substruct_ref.split ("_")[-2] substruct_parsed = parsePDB.loadCoordSectionPDB(p_substruct_ref, "HETATM") l_atom_substituate = neighborSearch.searchNeighborAtom(substruct_parsed, lig_parsed, struct_type, p_log, thresold_superimposed_ribose = thresold_superimposed_ribose, thresold_superimposed_pi = thresold_superimposed_pi) # control find if len (l_atom_substituate) == 0 : if not struct_type in d_control["subref empty"].keys () : d_control["subref empty"][struct_type] = 1 else : d_control["subref empty"][struct_type] = d_control["subref empty"][struct_type] + 1 continue else : if not struct_type in d_control["subref"].keys () : d_control["subref"][struct_type] = 1 else : d_control["subref"][struct_type] = d_control["subref"][struct_type] + 1 # write PDB file, convert smile p_substituate_pdb = p_dir_result_ref + "substituent_" + pdbfile.split ("_")[0] + "_" + pdbfile.split ("_")[1] + "_" + struct_type + ".pdb" writePDBfile.coordinateSection(p_substituate_pdb, l_atom_substituate, recorder="HETATM", header=0, connect_matrix = 1) # sheap reference on part of ligand p_sheap = runOtherSoft.runShaep (p_substruct_ref, p_substituate_pdb, p_substituate_pdb[0:-4] + ".hit", clean = 0) val_sheap = parseShaep.parseOutputShaep (p_sheap) if val_sheap == {} : p_log.write ("[ERROR] -> ShaEP " + p_substituate_pdb + " " + p_substruct_ref + "\n") if not struct_type in d_control["out sheap"].keys () : d_control["out sheap"][struct_type] = 1 else : d_control["out sheap"][struct_type] = d_control["out sheap"][struct_type] + 1 continue # control thresold sheap if not struct_type in d_filout_sheap.keys () : d_filout_sheap[struct_type] = {} d_filout_sheap[struct_type] = open (p_dir_result + "shaep_global_" + struct_type + ".txt", "w") d_filout_sheap[struct_type].write ("name\tbest_similarity\tshape_similarity\tESP_similarity\n") d_filout_sheap["list"].append (p_dir_result + "shaep_global_" + struct_type + ".txt") # to improve with python function # write value in ShaEP control d_filout_sheap[struct_type].write (ref_folder + "_" + str(pdbfile.split ("_")[1]) + "_" + struct_type + "_" + str (pdbfile.split ("_")[0]) + "\t" + str(val_sheap["best_similarity"]) + "\t" + str(val_sheap["shape_similarity"]) + "\t" + str(val_sheap["ESP_similarity"]) + "\n") d_filout_sheap["global"].write (ref_folder + "_" + str(pdbfile.split ("_")[1]) + "_" + struct_type + "_" + str (pdbfile.split ("_")[0]) + "\t" + str(val_sheap["best_similarity"]) + "\t" + str(val_sheap["shape_similarity"]) + "\t" + str(val_sheap["ESP_similarity"]) + "\n") # rename file substituent with shaEP value rename(p_substituate_pdb, p_substituate_pdb[:-4] + "_" + str (val_sheap["best_similarity"]) + ".pdb") # rename and change the file name p_substituate_pdb = p_substituate_pdb[:-4] + "_" + str (val_sheap["best_similarity"]) + ".pdb" # write all substruct in global file writePDBfile.coordinateSection(d_filout_superimposed["global"], lig_parsed, recorder= "HETATM", header = str(p_lig.split ("/")[-1]) + "_" + str (val_sheap["best_similarity"]) , connect_matrix = 1) # control sheap thresold if float(val_sheap["best_similarity"]) >= thresold_shaep : # write subligand superimposed selected in global files writePDBfile.coordinateSection(d_filout_superimposed["sheap"], lig_parsed, recorder= "HETATM", header = str(p_lig.split ("/")[-1]) + "_" + str (val_sheap["best_similarity"]) , connect_matrix = 1) ############ # write BS # ############ # not only protein superimposed -> also ion and water l_atom_complex = parsePDB.loadCoordSectionPDB(p_complex) superposeStructure.applyMatrixProt(l_atom_complex, p_matrix) p_file_cx = p_dir_result_ref + "CX_" + p_lig.split ("/")[-1] # write CX writePDBfile.coordinateSection(p_file_cx, l_atom_complex, recorder="ATOM", header= p_lig.split ("/")[-1], connect_matrix = 0) # search atom in BS l_atom_binding_site = [] for atom_complex in l_atom_complex : for atom_substruct in lig_parsed : if parsePDB.distanceTwoatoms (atom_substruct, atom_complex) <= thresold_BS : if not atom_complex in l_atom_binding_site : l_atom_binding_site.append (deepcopy(atom_complex)) # 3. retrieve complet residue l_atom_BS_res = parsePDB.getResidues(l_atom_binding_site, l_atom_complex) # 4. write binding site p_binding = p_dir_result_ref + "BS_" + p_lig.split ("/")[-1] writePDBfile.coordinateSection(p_binding, l_atom_BS_res, "ATOM", p_binding, connect_matrix = 0) # smile code substituate analysis # Step smile -> not conversion if shaep not validate smile_find = runOtherSoft.babelConvertPDBtoSMILE(p_substituate_pdb) if not struct_type in d_smile.keys () : d_smile[struct_type] = {} d_smile[struct_type][smile_find] = {} d_smile[struct_type][smile_find]["count"] = 1 d_smile[struct_type][smile_find]["PDB"] = [pdbfile.split ("_")[1]] d_smile[struct_type][smile_find]["ligand"] = [pdbfile.split ("_")[0]] d_smile[struct_type][smile_find]["ref"] = [ref_folder] else : if not smile_find in d_smile[struct_type].keys () : d_smile[struct_type][smile_find] = {} d_smile[struct_type][smile_find]["count"] = 1 d_smile[struct_type][smile_find]["PDB"] = [pdbfile.split ("_")[1]] d_smile[struct_type][smile_find]["ligand"] = [pdbfile.split ("_")[0]] d_smile[struct_type][smile_find]["ref"] = [ref_folder] else : d_smile[struct_type][smile_find]["count"] = d_smile[struct_type][smile_find]["count"] + 1 d_smile[struct_type][smile_find]["PDB"].append (pdbfile.split ("_")[1]) d_smile[struct_type][smile_find]["ligand"].append (pdbfile.split ("_")[0]) d_smile[struct_type][smile_find]["ref"].append (ref_folder) else : if not struct_type in d_control["out sheap"].keys () : d_control["out sheap"][struct_type] = 1 else : d_control["out sheap"][struct_type] = d_control["out sheap"][struct_type] + 1 tool.closeDicoFile (d_filout_superimposed) # sheap control tool.closeDicoFile (d_filout_sheap) for p_file_sheap in d_filout_sheap["list"] : runOtherSoft.RhistogramMultiple (p_file_sheap) # write list of smile for substruct in d_smile.keys () : p_list_smile = pathManage.result(name_lig) + "list_" + substruct + "_" + str (thresold_shaep) + "_smile.txt" filout_smile = open (p_list_smile, "w") for smile_code in d_smile[substruct].keys () : l_lig = d_smile[substruct][smile_code]["ligand"] l_PDB = d_smile[substruct][smile_code]["PDB"] l_ref = d_smile[substruct][smile_code]["ref"] filout_smile.write (str (smile_code) + "\t" + str (d_smile[substruct][smile_code]["count"]) + "\t" + " ".join (l_PDB) + "\t" + " ".join (l_ref) + "\t" + " ".join(l_lig) + "\n") filout_smile.close () p_log.close () # control filout_control.write ("NB ref: " + str(d_control["pr ref"]) + "\n") filout_control.write ("Ligand query: " + str(d_control["lig query"]) + "\n") for k in d_control["subref"].keys () : filout_control.write ("LSR " + str (k) + ": " + str(d_control["subref"][k]) + "\n") for k in d_control["subref empty"].keys () : filout_control.write ("NB LSR empty " + str (k) + ": " + str(d_control["subref empty"][k]) + "\n") for k in d_control["out sheap"].keys () : filout_control.write ("LSR out by sheap " + str (k) + ": " + str(d_control["out sheap"][k]) + "\n") filout_control.write ("**********************\n\n") for k in d_control["subref"].keys () : filout_control.write ("LSR keep" + str (k) + ": " + str(d_control["subref"][k] - d_control["out sheap"][k]) + "\n") filout_control.close () return 1
def analysisBS(name_lig, ID_seq='0.0', debug=1): pr_result = pathManage.result(name_lig) pr_out = pathManage.result(name_lig + "/sameBS") # log files p_log_file = pr_out + "log.txt" filout_log = open(p_log_file, "w") # dictionnar with files d_file_BS = {} d_file_BS["global"] = open(pr_out + name_lig + "_", "w") d_file_BS["global"].write( "name_bs\tRMSD_prot\tRMSD_BS_ca\tRMSD_BS_all\tD_max\tl_at_BS\tidentic\n" ) d_file_BS["summary"] = open(pr_out + "summary.txt", "w") pr_dataset = pathManage.dataset(name_lig) l_folder_ref = listdir(pr_result) nb_BS = 0 nb_BS_filtered = 0 nb_same_BS = 0 for PDB_ref in l_folder_ref: if debug: print PDB_ref if len(PDB_ref) != 4: continue p_pdb_ref = pathManage.findPDBRef(pr_dataset + PDB_ref + "/") l_p_query = pathManage.findPDBQueryTransloc( pathManage.result(name_lig) + PDB_ref + "/") if debug: print l_p_query for p_query in l_p_query: # read TM Align if debug: print p_query.split("/")[-1][7:-4] p_TMalign = pathManage.alignmentOutput(name_lig) + p_pdb_ref.split( "/")[-1][0:-4] + "__" + p_query.split("/")[-1][7:-4] + "/RMSD" try: score_align = parseTMalign.parseOutputTMalign(p_TMalign) except: filout_log.write("ERROR TM align " + p_TMalign + "\n") continue nb_BS = nb_BS + 1 if score_align["IDseq"] >= ID_seq: nb_BS_filtered = nb_BS_filtered + 1 l_p_substruct_ref = pathManage.findSubstructRef( pr_dataset + PDB_ref + "/", name_lig) # sub BS for p_substruct_ref in l_p_substruct_ref: struct_substitued = p_substruct_ref.split("_")[-2] # write header if not struct_substitued in d_file_BS.keys(): d_file_BS[struct_substitued] = open( pr_out + name_lig + "_" + struct_substitued + "_", "w") d_file_BS[struct_substitued].write( "name_bs\tRMSD_prot\tRMSD_BS_ca\tRMSD_BS_all\tD_max\tl_at_BS\tidentic\n" ) RMSD_bs = analysis.computeRMSDBS(p_pdb_ref, p_query, p_substruct_ref, pr_out) if RMSD_bs != []: d_file_BS[struct_substitued].write( p_substruct_ref.split("/")[-1][0:-4] + "_*_" + p_query.split("/")[-1][0:-4] + "\t" + str(score_align["RMSD"]) + "\t" + str(RMSD_bs[1]) + "\t" + str(RMSD_bs[0]) + "\t" + str(RMSD_bs[2]) + "\t" + str(RMSD_bs[-2]) + "\t" + str(RMSD_bs[-1]) + "\n") p_ligand_ref = pathManage.findligandRef( pr_dataset + PDB_ref + "/", name_lig) RMSD_bs_lig = analysis.computeRMSDBS(p_pdb_ref, p_query, p_ligand_ref, pr_out) if RMSD_bs_lig != []: d_file_BS["global"].write( p_ligand_ref.split("/")[-1][0:-4] + "_*_" + p_query.split("/")[-1][0:-4] + "\t" + str(score_align["RMSD"]) + "\t" + str(RMSD_bs_lig[1]) + "\t" + str(RMSD_bs_lig[0]) + "\t" + str(RMSD_bs_lig[2]) + "\t" + str(RMSD_bs_lig[-2]) + "\t" + str(RMSD_bs_lig[-1]) + "\n") if RMSD_bs_lig[-1] == 1: nb_same_BS = nb_same_BS + 1 # write summary d_file_BS["summary"].write("BS global: " + str(nb_BS) + "\n") d_file_BS["summary"].write("BS - IDseq " + str(ID_seq) + "%: " + str(nb_BS_filtered) + "\n") d_file_BS["summary"].write("BS - same atom number: " + str(nb_same_BS) + "\n") filout_log.close() # close files and run histograms for k_dico in d_file_BS.keys(): p_file = d_file_BS[k_dico].name d_file_BS[k_dico].close() if name_lig == "ATP": runOtherSoft.RhistogramRMSD(p_file, max_RMSD=5.0) elif name_lig == "ADP": runOtherSoft.RhistogramRMSD(p_file, max_RMSD=4.0) elif name_lig == "AMP": runOtherSoft.RhistogramRMSD(p_file, max_RMSD=4.0) else: runOtherSoft.RhistogramRMSD(p_file, max_RMSD=3.5) return 1
def globalArrangement (pr_orgin, p_smile, p_family, name_ligand, l_ligand_out): # print "--------" # print pr_orgin # print p_smile # print p_family # print name_ligand # print "--------" subst = p_smile.split ("_")[-3] filin = open (p_smile, "r") l_line_smile = filin.readlines () filin.close() for line_smile in l_line_smile : # search substructure # print line_smile l_PDB_query = line_smile.split ("\t")[-3].split (" ") # print l_PDB_query l_PDB_ref = line_smile.split ("\t")[-2].split (" ") l_ligand = line_smile.strip().split ("\t")[-1].split (" ") # search replacement smile = line_smile.split ("\t")[0] # search if LSR is small -> thresold < 3 small_LSR = smileAnalysis.smallLSR (smile) if subst == "ribose" : if small_LSR == 1 : first_folder = "ribose_small" else : first_folder = "ribose" else : if small_LSR == 1 : first_folder = "Pi_small" else : first_folder = "Pi" print smile, l_PDB_query, l_PDB_ref, l_ligand, subst, small_LSR replacement, metal = smileAnalysis.searchReplacement (smile, l_PDB_query[0], l_PDB_ref[0], name_ligand) # case with cycle -> search replacement 2 if replacement == "cycle" : replacement2, metal = smileAnalysis.searchReplacement (smile, l_PDB_query[0], l_PDB_ref[0], name_ligand, in_cycle = 1) replacement = replacement + "/" + replacement2 # new folder # case metal if replacement == "metal" : print metal, l_PDB_query, l_PDB_ref, name_ligand len_find = len (l_PDB_ref) i = 0 while i < len_find : # exclusion of ligand out if l_ligand[i] in l_ligand_out : i = i + 1 continue group, family = analysis.findFamilyAndGroup(l_PDB_ref[i]) # folder reference pr_dataset = pathManage.dataset(name_ligand + "/" + l_PDB_ref[i]) PDB_ref = pathManage.findPDBRef(pr_dataset) p_ligand_ref = pathManage.findligandRef(pr_dataset , name_ligand) l_frag_ref = pathManage.findSubstructRef(pr_dataset, name_ligand) for f_ref in l_frag_ref : if search (subst, f_ref) : p_frag_ref = f_ref break # folder_query pr_result = pathManage.result(name_ligand + "/" + l_PDB_ref[i]) l_protein_tranloc = pathManage.findPDBQueryTransloc(pr_result) for p_t in l_protein_tranloc : if search (l_ligand[i], p_t) and search (l_PDB_query[i], p_t) : p_protein_query = p_t break if replacement != "metal" : p_lig_query = pathManage.findligandQuery(pr_dataset , l_ligand[i], l_PDB_query[i]) else : p_lig_query = pathManage.findligandQuery(pr_dataset ,metal, l_PDB_query[i]) # need apply transloc matrix matrix_transloc = pathManage.findMatrix(p_ligand_ref, p_lig_query, name_ligand) lig_query_parsed = parsePDB.loadCoordSectionPDB(p_lig_query) try : superposeStructure.applyMatrixLigand(lig_query_parsed, matrix_transloc) except : i = i + 1 continue p_lig_substituate = pathManage.findSubstructFind(pr_result, l_ligand[i], l_PDB_query[i], subst) l_p_BS = pathManage.findFileBS(pr_result, l_PDB_query[i]) for BS in l_p_BS : if search (l_ligand[i], BS) : p_BS = BS break # print pr_final # print "***************" # print PDB_ref # print p_ligand_ref # print p_frag_ref # print "----" # print p_protein_query # print p_lig_query # print p_lig_substituate # print p_BS # print "**************" # ajouter group + family 2 lettre pr_final = pr_orgin + first_folder + "/" + replacement + "/" + str (family) + "-" + str (group) + "_" + l_PDB_ref[i] + "/" pr_ligand = pr_orgin + first_folder + "/" + replacement + "/" + str (family) + "-" + str (group) + "_" + l_PDB_ref[i] + "/LGD/" pr_BS = pr_orgin + first_folder + "/" + replacement + "/" + str (family) + "-" + str (group) + "_" + l_PDB_ref[i] + "/BS/" pr_sust = pr_orgin + first_folder + "/" + replacement + "/" + str (family) + "-" + str (group) + "_" + l_PDB_ref[i] + "/LSR/" if not path.isdir(pr_final): makedirs (pr_final) if not path.isdir(pr_ligand): makedirs (pr_ligand) if not path.isdir(pr_BS): makedirs (pr_BS) if not path.isdir(pr_sust): makedirs (pr_sust) # list file p_list_smile_queries = pr_sust + "list.smile" if not path.exists(p_list_smile_queries) : file_smile_queries = open (p_list_smile_queries, "w") else : file_smile_queries = open (p_list_smile_queries, "a") file_smile_queries.write (str(smile) + "\n") file_smile_queries.close () # lig de la query writePDBfile.coordinateSection(pr_ligand + "LGD_" + p_lig_query.split ("/")[-1], lig_query_parsed, recorder = "HETATM", header = "LCG_" + p_lig_query.split ("/")[-1], connect_matrix = 1) runOtherSoft.babelConvertPDBtoSMILE(pr_ligand + "LGD_" + p_lig_query.split ("/")[-1], clean_smi = 1) # lig de reference + smile copy2(p_ligand_ref, pr_ligand + "LGD_REF_" + p_ligand_ref.split ("/")[-1]) runOtherSoft.babelConvertPDBtoSMILE(pr_ligand + "LGD_REF_" + p_ligand_ref.split ("/")[-1]) # LSR de ref copy2(p_frag_ref, pr_sust + "LSR_REF_" + name_ligand + "_" + l_PDB_ref[i] + ".pdb") # protein query #copy2(p_protein_query, pr_final) # LSR query -> p_lig_ref only for the name copy2(p_lig_substituate, pr_sust + "LSR_" + subst + "_" + p_lig_query.split ("/")[-1]) # BS query copy2(p_BS, pr_BS) # BS from reference l_atom_BS = parsePDB.computeBS (PDB_ref, p_ligand_ref, thresold = 4.50, option_onlyATOM = 0) writePDBfile.coordinateSection(pr_BS + "BS_REF_" + name_ligand + "_" + PDB_ref.split ("/")[-1], l_atom_BS, recorder = "ATOM", header = "BS_REF_" + name_ligand + "_" + PDB_ref, connect_matrix = 0) i = i + 1 return 1
def retrieveSubstructSuperimposed(name_lig, thresold_BS=4.5, thresold_superimposed_ribose=2.5, thresold_superimposed_pi=3, thresold_shaep=0.4): # ouput p_dir_dataset = pathManage.dataset(name_lig) p_dir_result = pathManage.result(name_lig) l_folder_ref = listdir(p_dir_dataset) # log control p_log = open(p_dir_result + "log_superimposed.txt", "w") # control extraction d_control = {} d_control["pr ref"] = 0 d_control["lig query"] = 0 d_control["subref"] = {} d_control["subref empty"] = {} d_control["out sheap"] = {} filout_control = open(p_dir_result + "quality_extraction.txt", "w") # stock smile code d_smile = {} # sheap control d_filout_sheap = {} d_filout_sheap["list"] = [p_dir_result + "shaep_global.txt"] d_filout_sheap["global"] = open(p_dir_result + "shaep_global.txt", "w") d_filout_sheap["global"].write( "name\tbest_similarity\tshape_similarity\tESP_similarity\n") for ref_folder in l_folder_ref: # control folder reference name if len(ref_folder) != 4: p_log.write("[ERROR folder] -> " + ref_folder + "\n") continue # reference p_lig_ref = pathManage.findligandRef(p_dir_dataset + ref_folder + "/", name_lig) try: lig_ref_parsed = parsePDB.loadCoordSectionPDB(p_lig_ref, "HETATM") # print len (lig_ref_parsed) except: p_log.write("[ERROR ligand ref] -> " + p_lig_ref + "\n") continue #control d_control["pr ref"] = d_control["pr ref"] + 1 # output by reference p_dir_result_ref = pathManage.result(name_lig + "/" + ref_folder) d_filout_superimposed = {} d_filout_superimposed["global"] = open( p_dir_result_ref + "all_ligand_aligned.pdb", "w") d_filout_superimposed["sheap"] = open( p_dir_result_ref + "all_ligand_aligned_" + str(thresold_shaep) + ".pdb", "w") # write lig ref -> connect matrix corrrect in all reference and all sheap writePDBfile.coordinateSection(d_filout_superimposed["global"], lig_ref_parsed, "HETATM", connect_matrix=1) writePDBfile.coordinateSection(d_filout_superimposed["sheap"], lig_ref_parsed, "HETATM", connect_matrix=1) # inspect folder dataset l_pdbfile = listdir(p_dir_dataset + ref_folder + "/") for pdbfile in l_pdbfile: # no ligand file if len(pdbfile.split("_")) == 1: continue pdbfile = pdbfile[:-4] # remove extention if len(pdbfile.split("_")[0]) == 3 and len(pdbfile.split( "_")[1]) == 4 and pdbfile.split("_")[1] != ref_folder: p_lig = p_dir_dataset + ref_folder + "/" + pdbfile + ".pdb" if p_lig_ref != p_lig: # pass case where ligand replace same ligand -> does not need run if pdbfile.split("_")[0] == name_lig: p_log.write("[REMOVE] -> same ligand substituate") continue # parsed ligand query lig_parsed = parsePDB.loadCoordSectionPDB(p_lig, "HETATM") # find matrix of rotation p_matrix = pathManage.findMatrix(p_lig_ref, p_lig, name_lig) # control file matrix exist if not path.exists(p_matrix): p_log.write("[ERROR] -> Matrix transloc " + p_lig_ref + " " + p_lig + " " + name_lig + "\n") continue # control d_control["lig query"] = d_control["lig query"] + 1 # find the path of complex used p_complex = p_dir_dataset + ref_folder + "/" + p_lig.split( "/")[-1][4:] # ligand rotated -> change the referentiel superposeStructure.applyMatrixLigand(lig_parsed, p_matrix) # use substruct l_p_substruct_ref = pathManage.findSubstructRef( pathManage.dataset(name_lig) + ref_folder + "/", name_lig) for p_substruct_ref in l_p_substruct_ref: # ribose or phosphate struct_type = p_substruct_ref.split("_")[-2] substruct_parsed = parsePDB.loadCoordSectionPDB( p_substruct_ref, "HETATM") l_atom_substituate = neighborSearch.searchNeighborAtom( substruct_parsed, lig_parsed, struct_type, p_log, thresold_superimposed_ribose= thresold_superimposed_ribose, thresold_superimposed_pi=thresold_superimposed_pi) # control find if len(l_atom_substituate) == 0: if not struct_type in d_control[ "subref empty"].keys(): d_control["subref empty"][struct_type] = 1 else: d_control["subref empty"][ struct_type] = d_control["subref empty"][ struct_type] + 1 continue else: if not struct_type in d_control["subref"].keys(): d_control["subref"][struct_type] = 1 else: d_control["subref"][struct_type] = d_control[ "subref"][struct_type] + 1 # write PDB file, convert smile p_substituate_pdb = p_dir_result_ref + "substituent_" + pdbfile.split( "_")[0] + "_" + pdbfile.split( "_")[1] + "_" + struct_type + ".pdb" writePDBfile.coordinateSection(p_substituate_pdb, l_atom_substituate, recorder="HETATM", header=0, connect_matrix=1) # sheap reference on part of ligand p_sheap = runOtherSoft.runShaep( p_substruct_ref, p_substituate_pdb, p_substituate_pdb[0:-4] + ".hit", clean=0) val_sheap = parseShaep.parseOutputShaep(p_sheap) if val_sheap == {}: p_log.write("[ERROR] -> ShaEP " + p_substituate_pdb + " " + p_substruct_ref + "\n") if not struct_type in d_control[ "out sheap"].keys(): d_control["out sheap"][struct_type] = 1 else: d_control["out sheap"][ struct_type] = d_control["out sheap"][ struct_type] + 1 continue # control thresold sheap if not struct_type in d_filout_sheap.keys(): d_filout_sheap[struct_type] = {} d_filout_sheap[struct_type] = open( p_dir_result + "shaep_global_" + struct_type + ".txt", "w") d_filout_sheap[struct_type].write( "name\tbest_similarity\tshape_similarity\tESP_similarity\n" ) d_filout_sheap["list"].append( p_dir_result + "shaep_global_" + struct_type + ".txt") # to improve with python function # write value in ShaEP control d_filout_sheap[struct_type].write( ref_folder + "_" + str(pdbfile.split("_")[1]) + "_" + struct_type + "_" + str(pdbfile.split("_")[0]) + "\t" + str(val_sheap["best_similarity"]) + "\t" + str(val_sheap["shape_similarity"]) + "\t" + str(val_sheap["ESP_similarity"]) + "\n") d_filout_sheap["global"].write( ref_folder + "_" + str(pdbfile.split("_")[1]) + "_" + struct_type + "_" + str(pdbfile.split("_")[0]) + "\t" + str(val_sheap["best_similarity"]) + "\t" + str(val_sheap["shape_similarity"]) + "\t" + str(val_sheap["ESP_similarity"]) + "\n") # rename file substituent with shaEP value rename( p_substituate_pdb, p_substituate_pdb[:-4] + "_" + str(val_sheap["best_similarity"]) + ".pdb") # rename and change the file name p_substituate_pdb = p_substituate_pdb[:-4] + "_" + str( val_sheap["best_similarity"]) + ".pdb" # write all substruct in global file writePDBfile.coordinateSection( d_filout_superimposed["global"], lig_parsed, recorder="HETATM", header=str(p_lig.split("/")[-1]) + "_" + str(val_sheap["best_similarity"]), connect_matrix=1) # control sheap thresold if float(val_sheap["best_similarity"] ) >= thresold_shaep: # write subligand superimposed selected in global files writePDBfile.coordinateSection( d_filout_superimposed["sheap"], lig_parsed, recorder="HETATM", header=str(p_lig.split("/")[-1]) + "_" + str(val_sheap["best_similarity"]), connect_matrix=1) ############ # write BS # ############ # not only protein superimposed -> also ion and water l_atom_complex = parsePDB.loadCoordSectionPDB( p_complex) superposeStructure.applyMatrixProt( l_atom_complex, p_matrix) p_file_cx = p_dir_result_ref + "CX_" + p_lig.split( "/")[-1] # write CX writePDBfile.coordinateSection( p_file_cx, l_atom_complex, recorder="ATOM", header=p_lig.split("/")[-1], connect_matrix=0) # search atom in BS l_atom_binding_site = [] for atom_complex in l_atom_complex: for atom_substruct in lig_parsed: if parsePDB.distanceTwoatoms( atom_substruct, atom_complex) <= thresold_BS: if not atom_complex in l_atom_binding_site: l_atom_binding_site.append( deepcopy(atom_complex)) # 3. retrieve complet residue l_atom_BS_res = parsePDB.getResidues( l_atom_binding_site, l_atom_complex) # 4. write binding site p_binding = p_dir_result_ref + "BS_" + p_lig.split( "/")[-1] writePDBfile.coordinateSection( p_binding, l_atom_BS_res, "ATOM", p_binding, connect_matrix=0) # smile code substituate analysis # Step smile -> not conversion if shaep not validate smile_find = runOtherSoft.babelConvertPDBtoSMILE( p_substituate_pdb) if not struct_type in d_smile.keys(): d_smile[struct_type] = {} d_smile[struct_type][smile_find] = {} d_smile[struct_type][smile_find][ "count"] = 1 d_smile[struct_type][smile_find]["PDB"] = [ pdbfile.split("_")[1] ] d_smile[struct_type][smile_find][ "ligand"] = [pdbfile.split("_")[0]] d_smile[struct_type][smile_find]["ref"] = [ ref_folder ] else: if not smile_find in d_smile[ struct_type].keys(): d_smile[struct_type][smile_find] = {} d_smile[struct_type][smile_find][ "count"] = 1 d_smile[struct_type][smile_find][ "PDB"] = [pdbfile.split("_")[1]] d_smile[struct_type][smile_find][ "ligand"] = [ pdbfile.split("_")[0] ] d_smile[struct_type][smile_find][ "ref"] = [ref_folder] else: d_smile[struct_type][smile_find][ "count"] = d_smile[struct_type][ smile_find]["count"] + 1 d_smile[struct_type][smile_find][ "PDB"].append( pdbfile.split("_")[1]) d_smile[struct_type][smile_find][ "ligand"].append( pdbfile.split("_")[0]) d_smile[struct_type][smile_find][ "ref"].append(ref_folder) else: if not struct_type in d_control[ "out sheap"].keys(): d_control["out sheap"][struct_type] = 1 else: d_control["out sheap"][ struct_type] = d_control["out sheap"][ struct_type] + 1 tool.closeDicoFile(d_filout_superimposed) # sheap control tool.closeDicoFile(d_filout_sheap) for p_file_sheap in d_filout_sheap["list"]: runOtherSoft.RhistogramMultiple(p_file_sheap) # write list of smile for substruct in d_smile.keys(): p_list_smile = pathManage.result( name_lig) + "list_" + substruct + "_" + str( thresold_shaep) + "_smile.txt" filout_smile = open(p_list_smile, "w") for smile_code in d_smile[substruct].keys(): l_lig = d_smile[substruct][smile_code]["ligand"] l_PDB = d_smile[substruct][smile_code]["PDB"] l_ref = d_smile[substruct][smile_code]["ref"] filout_smile.write( str(smile_code) + "\t" + str(d_smile[substruct][smile_code]["count"]) + "\t" + " ".join(l_PDB) + "\t" + " ".join(l_ref) + "\t" + " ".join(l_lig) + "\n") filout_smile.close() p_log.close() # control filout_control.write("NB ref: " + str(d_control["pr ref"]) + "\n") filout_control.write("Ligand query: " + str(d_control["lig query"]) + "\n") for k in d_control["subref"].keys(): filout_control.write("LSR " + str(k) + ": " + str(d_control["subref"][k]) + "\n") for k in d_control["subref empty"].keys(): filout_control.write("NB LSR empty " + str(k) + ": " + str(d_control["subref empty"][k]) + "\n") for k in d_control["out sheap"].keys(): filout_control.write("LSR out by sheap " + str(k) + ": " + str(d_control["out sheap"][k]) + "\n") filout_control.write("**********************\n\n") for k in d_control["subref"].keys(): filout_control.write("LSR keep" + str(k) + ": " + str(d_control["subref"][k] - d_control["out sheap"][k]) + "\n") filout_control.close() return 1