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 extractReference(p_list_ligand, p_dir_dataset, p_dir_result, substruct):
# struct reference
d_dataset = {}
# retrieve list of ligand in PDB
d_ligand = tool.parseLigandPDBList(p_list_ligand)
# download PDB and fasta associated
l_p_PDB = []
l_p_fasta = []
for PDB_ID in d_ligand[substruct]:
PDB_ID = PDB_ID.upper()
p_pdb = downloadFile.importPDB(PDB_ID,
p_dir_dataset,
dir_by_PDB=1,
debug=1,
dbPDB="/home/borrel/PDB/")
p_fasta = downloadFile.importFasta(
PDB_ID,
p_dir_dataset,
dir_by_PDB=1,
debug=1,
fastaGlobal="/home/borrel/Yue_project/pdb_seqres.txt")
if p_pdb != 0 and p_fasta != 0:
l_p_pdb_chain = separeByChain(p_pdb)
l_p_fasta_chain = separeChainFasta(p_fasta)
d_dataset[PDB_ID] = {}
d_dataset[PDB_ID]["p_pdb"] = p_pdb
d_dataset[PDB_ID]["p_fasta"] = p_fasta
d_dataset[PDB_ID]["p_pdb_chain"] = l_p_pdb_chain
d_dataset[PDB_ID]["p_fasta_chain"] = l_p_fasta_chain
d_dataset[PDB_ID]["conserve"] = 1
# plot resolution
p_file_RX = p_dir_result + "resolution_ref.txt"
file_RX = open(p_file_RX, "w")
for PDB_ID in d_dataset.keys():
RX = parsePDB.resolution(d_dataset[PDB_ID]["p_pdb"])
try:
d_dataset[PDB_ID]["RX"] = float(RX)
except:
d_dataset[PDB_ID]["RX"] = 100.0
file_RX.write(PDB_ID + "\t" + str(RX) + "\n")
file_RX.close()
runOtherSoft.Rhistogram(p_file_RX, "RX_ref_no_filter")
return d_dataset
def extractReference (p_list_ligand, p_dir_dataset, p_dir_result, substruct):
# struct reference
d_dataset = {}
# retrieve list of ligand in PDB
d_ligand = tool.parseLigandPDBList (p_list_ligand)
# download PDB and fasta associated
l_p_PDB = []
l_p_fasta = []
for PDB_ID in d_ligand[substruct] :
PDB_ID = PDB_ID.upper()
p_pdb = downloadFile.importPDB(PDB_ID, p_dir_dataset, dir_by_PDB = 1, debug = 1, dbPDB = "/home/borrel/PDB/" )
p_fasta = downloadFile.importFasta(PDB_ID, p_dir_dataset, dir_by_PDB = 1, debug = 1, fastaGlobal = "/home/borrel/Yue_project/pdb_seqres.txt")
if p_pdb != 0 and p_fasta != 0 :
l_p_pdb_chain = separeByChain (p_pdb)
l_p_fasta_chain = separeChainFasta(p_fasta)
d_dataset[PDB_ID] = {}
d_dataset[PDB_ID] ["p_pdb"] = p_pdb
d_dataset[PDB_ID] ["p_fasta"] = p_fasta
d_dataset[PDB_ID] ["p_pdb_chain"] = l_p_pdb_chain
d_dataset[PDB_ID] ["p_fasta_chain"] = l_p_fasta_chain
d_dataset[PDB_ID] ["conserve"] = 1
# plot resolution
p_file_RX = p_dir_result + "resolution_ref.txt"
file_RX = open (p_file_RX, "w")
for PDB_ID in d_dataset.keys () :
RX = parsePDB.resolution(d_dataset[PDB_ID]["p_pdb"])
try : d_dataset[PDB_ID] ["RX"] = float(RX)
except : d_dataset[PDB_ID] ["RX"] = 100.0
file_RX.write (PDB_ID + "\t" + str (RX) + "\n")
file_RX.close ()
runOtherSoft.Rhistogram (p_file_RX, "RX_ref_no_filter")
return d_dataset
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