def selectBestFilePDBFasta (d_PDB, result_align, substruct) :
# case with only one PDB
d_PDB["best"] = {}
if result_align == [] :
d_PDB["best"]["PDB"] = d_PDB["p_pdb"]
d_PDB["best"]["fasta"] = d_PDB["p_fasta"]
else :
for primary_key in result_align.keys () :
for secondary_key in result_align[primary_key] .keys ():
if result_align[primary_key][secondary_key] != "100.0%" :
d_PDB["best"]["PDB"] = d_PDB["p_pdb"]
d_PDB["best"]["fasta"] = d_PDB["p_fasta"]
# fusion fasta for blast remove header
tool.fusionchainfasta(d_PDB["p_fasta"])
return
# control substructure presented // case identity 100%
nb_chain = len (d_PDB["p_pdb_chain"])
i = 0
while i < nb_chain :
l_ligand = parsePDB.retrieveListLigand(d_PDB["p_pdb_chain"][i])
if substruct in l_ligand :
# print "999999",d_PDB
d_PDB["best"]["PDB"] = d_PDB["p_pdb_chain"][i]
d_PDB["best"]["fasta"] = d_PDB["p_fasta_chain"][i]
return
else :
i = i + 1
def selectBestFilePDBFasta(d_PDB, result_align, substruct):
# case with only one PDB
d_PDB["best"] = {}
if result_align == []:
d_PDB["best"]["PDB"] = d_PDB["p_pdb"]
d_PDB["best"]["fasta"] = d_PDB["p_fasta"]
else:
for primary_key in result_align.keys():
for secondary_key in result_align[primary_key].keys():
if result_align[primary_key][secondary_key] != "100.0%":
d_PDB["best"]["PDB"] = d_PDB["p_pdb"]
d_PDB["best"]["fasta"] = d_PDB["p_fasta"]
# fusion fasta for blast remove header
tool.fusionchainfasta(d_PDB["p_fasta"])
return
# control substructure presented // case identity 100%
nb_chain = len(d_PDB["p_pdb_chain"])
i = 0
while i < nb_chain:
l_ligand = parsePDB.retrieveListLigand(d_PDB["p_pdb_chain"][i])
if substruct in l_ligand:
# print "999999",d_PDB
d_PDB["best"]["PDB"] = d_PDB["p_pdb_chain"][i]
d_PDB["best"]["fasta"] = d_PDB["p_fasta_chain"][i]
return
else:
i = i + 1
def datasetPreparation (ligand_ID, clean = 1):
p_dir_dataset = pathManage.dataset(ligand_ID)
l_folder = listdir(p_dir_dataset)
indent = 0
for ref_folder in l_folder :
# file include in dataset folder
if len (ref_folder) != 4:
continue
l_pdbfile = listdir(p_dir_dataset + ref_folder + "/")
indent = indent + 1
print ref_folder, indent
# clean repertory -> only PDB ref and PDB
l_pdbfile = listdir(p_dir_dataset + ref_folder + "/")
if clean == 1 :
for pdbfile in l_pdbfile :
p_file_pdb = p_dir_dataset + ref_folder + "/" + pdbfile
if not search (".pdb", pdbfile ) or search ("subref", pdbfile) or len (pdbfile.split("_")[0]) == 3:
remove (p_file_pdb)
l_pdbfile = listdir(p_dir_dataset + ref_folder + "/")
for pdbfile in l_pdbfile :
p_file_pdb = p_dir_dataset + ref_folder + "/" + pdbfile
# extract ligand in PDB
l_ligand = parsePDB.retrieveListLigand(p_file_pdb)
# print l_ligand
if l_ligand == []:
continue
else:
l_atom_pdb_parsed = parsePDB.loadCoordSectionPDB(p_file_pdb)
for name_ligand in l_ligand :
l_lig_parsed = parsePDB.retrieveLigand(l_atom_pdb_parsed, name_ligand)
if l_lig_parsed == [] :
continue
p_filout_ligand = p_dir_dataset + ref_folder + "/" + name_ligand + "_" + path.split(p_file_pdb)[1]
writePDBfile.coordinateSection(p_filout_ligand , l_lig_parsed[0], "HETATM", header=0 , connect_matrix = 1)
# ligand_ID write for shaep
# print p_dir_dataset + ref_folder + "/"
p_lig_ref = pathManage.findligandRef(p_dir_dataset + ref_folder + "/", ligand_ID)
if p_lig_ref == 0:
continue
# print p_lig_ref
lig_ref_parsed = parsePDB.loadCoordSectionPDB(p_lig_ref)
d_l_atom_substruct = substructTools.retrieveSubstruct(lig_ref_parsed, ligand_ID)
# case with AMP without phosphate
if d_l_atom_substruct == {}:
continue
# write ligand_ID
for subs in d_l_atom_substruct.keys ():
p_filout_substruct = p_dir_dataset + ref_folder + "/subref_" + subs + "_" + ref_folder + ".pdb"
writePDBfile.coordinateSection(p_filout_substruct , d_l_atom_substruct [subs], "HETATM", header=0 , connect_matrix = 1)
return 1
def filterBlastResult (d_dataset, p_dir_dataset, substruct, thresold_RX = 2.5, thresold_blast = 1e-4, debug = 1) :
"""
Filter resolution PDB
Filter evalue
"""
for pdb_ref in d_dataset.keys () :
if d_dataset[pdb_ref]["conserve"] == 0 : continue
for pdb_blast_chain in d_dataset[pdb_ref]["align"].keys () :
# filter e.value
if debug == 1 : print d_dataset[pdb_ref]["align"][pdb_blast_chain], pdb_ref, thresold_blast
# remove thresold and reference cleanner, remove if alignement on same protein
if d_dataset[pdb_ref]["align"][pdb_blast_chain] <= thresold_blast and pdb_blast_chain [0:4].upper() != pdb_ref:
if debug == 1 :print "CONTROL", pdb_blast_chain, d_dataset[pdb_ref]["align"][pdb_blast_chain]
# dowload PDB files
pdb_blast = pdb_blast_chain[0:4]
p_pdb_blast = downloadFile.importPDB(pdb_blast, p_dir_dataset + pdb_ref + "/", dir_by_PDB=0, dbPDB = "/home/borrel/PDB/")
if p_pdb_blast == 0 :
continue
# print p_pdb_blast, pdb_blast, "************"
l_queries_by_chain = separeByChain(p_pdb_blast) # divise chain in reference folder
try : RX = parsePDB.resolution(p_pdb_blast)
except : RX = 100.0
l_ligand = parsePDB.retrieveListLigand(p_pdb_blast)
if debug == 1 : print p_pdb_blast, l_ligand, RX, "control blast\n" + " ".join(l_queries_by_chain) + "\n"
# remove apo forms and remove not substiuant + list ligand do not considered
if l_ligand == [] or substruct in l_ligand:
remove (p_pdb_blast)
for queries_by_chain in l_queries_by_chain :
remove (queries_by_chain)
continue
# case RMN structure
try : RX = float(RX)
except: continue
if debug == 1 :print "----", RX, thresold_RX, "----"
if float(RX) <= thresold_RX :
if not "blast" in d_dataset[pdb_ref].keys () :
d_dataset[pdb_ref]["blast"] = [pdb_blast_chain.upper ()]
else :
# control if other chain in list
flag_in = 0
for PDB_in in d_dataset[pdb_ref]["blast"] :
if PDB_in.split ("_")[0] == pdb_blast_chain.upper ().split ("_")[0] :
flag_in = 1
if flag_in == 0 :
d_dataset[pdb_ref]["blast"].append (pdb_blast_chain.upper ())
if debug == 1 :print d_dataset[pdb_ref]["blast"], pdb_ref
def searchLigands(prPDB, prresult):
'''search ligands in PDB database
out : list of ligands with PDB files associated'''
print "Start Search Ligand In PDB file"
pfilout = prresult + "resultLigandInPDB"
# control file exist
if path.exists(pfilout) and path.getsize(pfilout) != 0:
return pfilout
l_PDB = retriveListPDB(prPDB)
filout = open(pfilout, "w")
for PDBid in l_PDB:
llig = parsePDB.retrieveListLigand(prPDB + PDBid.lower() + ".pdb")
if llig != []:
filout.write(PDBid + "\t" + " ".join(llig) + "\n")
else:
continue
filout.close()
return pfilout
def searchLigands(prPDB, prresult):
'''search ligands in PDB database
out : list of ligands with PDB files associated'''
print "Start Search Ligand In PDB file"
pfilout = prresult + "resultLigandInPDB"
# control file exist
if path.exists(pfilout) and path.getsize(pfilout) != 0:
return pfilout
l_PDB = retriveListPDB(prPDB)
filout = open(pfilout, "w")
for PDBid in l_PDB:
llig = parsePDB.retrieveListLigand(prPDB + PDBid.lower() + ".pdb")
if llig != []:
filout.write(PDBid + "\t" + " ".join(llig) + "\n")
else:
continue
filout.close()
return pfilout
def datasetPreparation(ligand_ID, clean=1):
p_dir_dataset = pathManage.dataset(ligand_ID)
l_folder = listdir(p_dir_dataset)
indent = 0
for ref_folder in l_folder:
# file include in dataset folder
if len(ref_folder) != 4:
continue
l_pdbfile = listdir(p_dir_dataset + ref_folder + "/")
indent = indent + 1
print ref_folder, indent
# clean repertory -> only PDB ref and PDB
l_pdbfile = listdir(p_dir_dataset + ref_folder + "/")
if clean == 1:
for pdbfile in l_pdbfile:
p_file_pdb = p_dir_dataset + ref_folder + "/" + pdbfile
if not search(".pdb", pdbfile) or search(
"subref", pdbfile) or len(pdbfile.split("_")[0]) == 3:
remove(p_file_pdb)
l_pdbfile = listdir(p_dir_dataset + ref_folder + "/")
for pdbfile in l_pdbfile:
p_file_pdb = p_dir_dataset + ref_folder + "/" + pdbfile
# extract ligand in PDB
l_ligand = parsePDB.retrieveListLigand(p_file_pdb)
# print l_ligand
if l_ligand == []:
continue
else:
l_atom_pdb_parsed = parsePDB.loadCoordSectionPDB(p_file_pdb)
for name_ligand in l_ligand:
l_lig_parsed = parsePDB.retrieveLigand(
l_atom_pdb_parsed, name_ligand)
if l_lig_parsed == []:
continue
p_filout_ligand = p_dir_dataset + ref_folder + "/" + name_ligand + "_" + path.split(
p_file_pdb)[1]
writePDBfile.coordinateSection(p_filout_ligand,
l_lig_parsed[0],
"HETATM",
header=0,
connect_matrix=1)
# ligand_ID write for shaep
# print p_dir_dataset + ref_folder + "/"
p_lig_ref = pathManage.findligandRef(p_dir_dataset + ref_folder + "/",
ligand_ID)
if p_lig_ref == 0:
continue
# print p_lig_ref
lig_ref_parsed = parsePDB.loadCoordSectionPDB(p_lig_ref)
d_l_atom_substruct = substructTools.retrieveSubstruct(
lig_ref_parsed, ligand_ID)
# case with AMP without phosphate
if d_l_atom_substruct == {}:
continue
# write ligand_ID
for subs in d_l_atom_substruct.keys():
p_filout_substruct = p_dir_dataset + ref_folder + "/subref_" + subs + "_" + ref_folder + ".pdb"
writePDBfile.coordinateSection(p_filout_substruct,
d_l_atom_substruct[subs],
"HETATM",
header=0,
connect_matrix=1)
return 1
def filterBlastResult(d_dataset,
p_dir_dataset,
substruct,
thresold_RX=2.5,
thresold_blast=1e-4,
debug=1):
"""
Filter resolution PDB
Filter evalue
"""
for pdb_ref in d_dataset.keys():
if d_dataset[pdb_ref]["conserve"] == 0: continue
for pdb_blast_chain in d_dataset[pdb_ref]["align"].keys():
# filter e.value
if debug == 1:
print d_dataset[pdb_ref]["align"][
pdb_blast_chain], pdb_ref, thresold_blast
# remove thresold and reference cleanner, remove if alignement on same protein
if d_dataset[pdb_ref]["align"][
pdb_blast_chain] <= thresold_blast and pdb_blast_chain[
0:4].upper() != pdb_ref:
if debug == 1:
print "CONTROL", pdb_blast_chain, d_dataset[pdb_ref][
"align"][pdb_blast_chain]
# dowload PDB files
pdb_blast = pdb_blast_chain[0:4]
p_pdb_blast = downloadFile.importPDB(pdb_blast,
p_dir_dataset + pdb_ref +
"/",
dir_by_PDB=0,
dbPDB="/home/borrel/PDB/")
if p_pdb_blast == 0:
continue
# print p_pdb_blast, pdb_blast, "************"
l_queries_by_chain = separeByChain(
p_pdb_blast) # divise chain in reference folder
try:
RX = parsePDB.resolution(p_pdb_blast)
except:
RX = 100.0
l_ligand = parsePDB.retrieveListLigand(p_pdb_blast)
if debug == 1:
print p_pdb_blast, l_ligand, RX, "control blast\n" + " ".join(
l_queries_by_chain) + "\n"
# remove apo forms and remove not substiuant + list ligand do not considered
if l_ligand == [] or substruct in l_ligand:
remove(p_pdb_blast)
for queries_by_chain in l_queries_by_chain:
remove(queries_by_chain)
continue
# case RMN structure
try:
RX = float(RX)
except:
continue
if debug == 1: print "----", RX, thresold_RX, "----"
if float(RX) <= thresold_RX:
if not "blast" in d_dataset[pdb_ref].keys():
d_dataset[pdb_ref]["blast"] = [pdb_blast_chain.upper()]
else:
# control if other chain in list
flag_in = 0
for PDB_in in d_dataset[pdb_ref]["blast"]:
if PDB_in.split("_")[0] == pdb_blast_chain.upper(
).split("_")[0]:
flag_in = 1
if flag_in == 0:
d_dataset[pdb_ref]["blast"].append(
pdb_blast_chain.upper())
if debug == 1: print d_dataset[pdb_ref]["blast"], pdb_ref