def SearchChemicalSubstruct (pr_data, pr_result, control = 0):
# substructure search
p_filout = pr_result + "findSruct"
filout = open (p_filout, "w")
l_ligand = listdir(pr_data)
print l_ligand
for ligand in l_ligand :
print ligand
l_file = listdir(pr_data + ligand + "/")
for f in l_file :
if ligand == "ZM241385" or ligand == "source":
group = f
else :
group = f.split ("_")[-2]
p_file_PDB = pr_data + ligand + "/" + f
print p_file_PDB
if control == 1 :
ControlPDBFormat(p_file_PDB)
l_atom = parsing.loadCoordSectionPDB(p_file_PDB, remove_H = 1)
if ligand == "ZM241385" :
ll_atom_lig = parsing.retrieveLigand(l_atom, "ZMA")
else :
ll_atom_lig = parsing.retrieveLigand(l_atom, "RES")
for l_atom_lig in ll_atom_lig :
l_subs = searchPDB.interestStructure(l_atom_lig, more_flex = 1)
filout.write (str (ligand) + "\t" + str (f) + "\t" + str (group) + "\t" + " ".join(l_subs) + "\n")
filout.close ()
# best group ergomine -> 2128 (2013) && 6882 (2013)
# best group taladegid -> 7527 (2013) Vanderblit
# best group eticlopride -> 1285 (2010)
# ZM241385 => mod7msp
def CheckComplexQuality(l_in, name_lig, limit_RX, limit_RFree, one_PDB_out, debug = 1):
"""Check if the PDB file is similar and preserve file with the best Quality
remove file contain only DNA, RNA structure
in : list of PDB files
out : void -> change directly PDB list"""
l_PDB = deepcopy(l_in)
nb_PDB = len(l_PDB)
l_RX = []
i = 0
while i < nb_PDB:
# load PDB
d_PDB_load = loadFile.ExtractInfoPDBID(l_PDB[i])
# case PDB not found => often difference between list and database
if d_PDB_load == {} :
del l_PDB[i]
nb_PDB = nb_PDB - 1
continue
# check Header
header_PDB = d_PDB_load["Header"]
# Check DNA or RNA
if search ("dna", header_PDB) or search ("rna", header_PDB):
if debug == 1 : print "Exit => out dna", l_PDB[i]
del l_PDB[i]
nb_PDB = nb_PDB - 1
continue
# check quality
RX = d_PDB_load["RX"]
RFree = d_PDB_load["RFree"]
if float (RX) > limit_RX or float (RFree) > limit_RFree :
if debug == 1 : print "Exit => Quality structure", l_PDB[i], RX, RFree
del l_PDB[i]
nb_PDB = nb_PDB - 1
continue
# check if the structure contain the substructure
# -> because when we improve the resolution the ligand change and the search PDB -> ERROR
l_interest_sub = searchPDB.interestStructure(d_PDB_load[name_lig][0])
if l_interest_sub == [] :
if debug == 1 : print "Exit => No Sub found", l_PDB[i], RX, RFree
del l_PDB[i]
nb_PDB = nb_PDB - 1
continue
# Check false ligand, hooked to the protein
l_atom_protein = d_PDB_load["protein"]
l_atom_lig = d_PDB_load[name_lig][0]
if parsing.checkLigandHooked (l_atom_protein, l_atom_lig) == 1:
if debug == 1 : print "Exit => ligand = Modificated Amino acid", l_PDB[i]
nb_PDB = nb_PDB - 1
del l_PDB[i]
continue
l_RX.append (RX)
i = i + 1
if len (l_PDB) == 1 or len (l_PDB) == 0 :
return l_PDB
# case where we want only one PDB -> retrieve best quality
if one_PDB_out == 1 :
return [l_PDB[l_PDB.index (min (l_PDB))]]
return l_PDB
def Builder(name_database, RX = 3.00, RFree = 0.25, one_PDB_by_lig = 0, debug = 1):
"""
Dataset Builder
in : - open file result of filter ligand PDB
out : - log file
- dataset file -> ligand with associated PDB
"""
if one_PDB_by_lig == 0 :
name_dataset = name_database + "/" + str (RX) + "_" + str (RFree) + "_multiPDB"
else :
name_dataset = name_database + "/" + str (RX) + "_" + str (RFree) + "_uniquePDB"
pr_database = pathManage.result(name_database)
pr_result = pathManage.result(name_dataset)
if debug : print "== Path result " + pr_result + "==\n"
# check dataSet exist !!!!!!
# short cut
l_file_dataset = pathManage.retriveDataSetFile (pr_result)
if len(l_file_dataset) != 0 :
return l_file_dataset
# load structure
d_lig_PDB = loadFile.LigandInPDB(pr_database + "resultLigandInPDB")
nb_lig = len(d_lig_PDB.keys())
print "NB ligand included database:", nb_lig
# print d_lig_PDB.keys().index("HSO") -> search index ligand
i = 0
while i < nb_lig:
name_lig = d_lig_PDB.keys()[i]
#######################################
# step 1 search chemical substructure #
#######################################
PDB_ref = d_lig_PDB[name_lig][0]
if debug : print PDB_ref, name_lig, i, nb_lig
# if not possible to load the ligand -> remove lig
try : l_atom_lig_ref = loadFile.ligandInPDBConnectMatrixLigand(PDB_ref, name_lig)
except :
if debug == 1 : print "Exit => load ligand-l59"
del d_lig_PDB[name_lig]
nb_lig = nb_lig - 1
continue
# search substructure interest
l_interest_sub = searchPDB.interestStructure(l_atom_lig_ref) # search interest structure
if debug : print "Interest substructures in " + str(name_lig) + "-" + str (PDB_ref) + " " + "-".join(l_interest_sub)
if l_interest_sub == []:
if debug == 1 : print "Exit => Not substructure-l68"
del d_lig_PDB[name_lig]
nb_lig = nb_lig - 1
continue
#######################################################
# Step 2 Control quality of PDB + ligand hooked + option one #
#######################################################
else :
# control dataset quality
if debug : print "List PDB checked -> ", d_lig_PDB[name_lig]
l_PDB = checkPDBfile.CheckComplexQuality(d_lig_PDB[name_lig], name_lig, RX, RFree, one_PDB_by_lig)
# remove the entrance key with the ligand
if l_PDB == []:
if debug == 1 : print "Exit => Not No PDB selected-l82"
del d_lig_PDB[name_lig]
nb_lig = nb_lig - 1
continue
else :
d_lig_PDB[name_lig] = l_PDB
i = i + 1
if debug == 1 : print "Number of ligand selected =>", nb_lig
# structure and file dataset and control RX + length bond
WriteDataset (d_lig_PDB, pr_result)
return Builder(name_database, RX , RFree , one_PDB_by_lig , debug = 1)