def countingSubstituent (name_final, debug = 1):
pr_final_folder = pathManage.result("final_" + name_final)
d_count = {}
d_lig = {}
d_by_ref = {}
d_count_pr = {}
l_file_final = listdir(pr_final_folder)
if debug : print "1", pr_final_folder
for pr_type_subref in l_file_final :
# case where pr type is a file not a folder
try : l_pr_sub = listdir(pr_final_folder + pr_type_subref + "/")
except : continue
if debug: print "2",pr_final_folder + pr_type_subref + "/"
# case cycle append one directory
if "cycle" in l_pr_sub :
l_pr_sub.remove ("cycle")
l_second_sub = listdir (pr_final_folder + pr_type_subref + "/cycle/")
for second_sub in l_second_sub :
l_pr_sub.append ("cycle/" + second_sub)
for pr_sub in l_pr_sub :
# case where pr_type_substituent is a folder
try : l_pr_PDBref = listdir(pr_final_folder + pr_type_subref + "/" + pr_sub + "/")
except : continue
if debug : print "3", pr_final_folder + pr_type_subref, pr_sub
for pr_PDBref in l_pr_PDBref :
PDB_ref = pr_PDBref.split ("_")[-1]
family_ref = pr_PDBref.split ("-")[0]
group_ref = pr_PDBref.split ("_")[0].split ("-")[-1]
pr_LGD = pr_final_folder + pr_type_subref + "/" + pr_sub + "/" + pr_PDBref + "/LGD/"
pr_LSR = pr_final_folder + pr_type_subref + "/" + pr_sub + "/" + pr_PDBref + "/LSR/"
pr_BS = pr_final_folder + pr_type_subref + "/" + pr_sub + "/" + pr_PDBref + "/BS/"
if debug :
print "4",pr_LGD
print "4", pr_BS
print "4", pr_LSR
################
# folder LSR #
################
l_file_LSR = listdir (pr_LSR)
for file_LSR in l_file_LSR :
# -> count by type sub reference
if search ("LSR_", file_LSR) and file_LSR.split ("_")[1] != "REF" :
ligand_sub = file_LSR.split ("_")[1]
if debug : print "5", file_LSR
if not ligand_sub in d_count.keys () :
d_count[ligand_sub] = {}
if not pr_sub in d_count[ligand_sub].keys () :
d_count[ligand_sub][pr_sub] = 0
d_count[ligand_sub][pr_sub] = d_count[ligand_sub][pr_sub] + 1
################
# complet LSR #
################
elif search ("LSR", file_LSR):
# case LSR reference #
######################
if search ("REF_", file_LSR) :
lig_ref = file_LSR.split ("_")[2][:3]
if not lig_ref in d_by_ref.keys () :
d_by_ref[lig_ref] = {}
type_ref = pr_type_subref.split ("_")[0]
if not type_ref in d_by_ref[lig_ref].keys () :
d_by_ref[lig_ref][type_ref] = 0
d_by_ref[lig_ref][type_ref] = d_by_ref[lig_ref][type_ref] + 1
#################
# folder LGD #
#################
l_file_LGD = listdir(pr_LGD)
for file_LGD in l_file_LGD :
# print file_ref
if search ("LGD", file_LGD):
ligand = file_LGD.split ("_")[1]
if ligand == "REF" :
continue
if not ligand in d_lig.keys () :
d_lig[ligand] = {}
d_lig[ligand]["count"] = 0
d_lig[ligand]["group"] = []
d_lig[ligand]["family"] = []
d_lig[ligand]["count"] = d_lig[ligand]["count"] + 1
d_lig[ligand]["family"].append (str(family_ref))
d_lig[ligand]["group"].append (str(group_ref))
###############
# folder BS #
###############
l_file_BS = listdir(pr_BS)
for file_BS in l_file_BS :
if search ("BS_REF", file_BS):
lig_ref = file_BS.split ("_")[2]
pr_ref = file_BS.split ("_")[3].split (".")[0]
print lig_ref, pr_ref, "*****"
if not lig_ref in d_count_pr.keys () :
d_count_pr[lig_ref] = {}
d_count_pr[lig_ref]["pr ref"] = []
d_count_pr[lig_ref]["pr queries"] = []
d_count_pr[lig_ref]["lig queries"] = []
if not pr_ref in d_count_pr[lig_ref]["pr ref"] :
d_count_pr[lig_ref]["pr ref"].append (pr_ref)
try:
family = analysis.findFamily (pr_ref, pathManage.dataset (lig_ref) + "family_PDB.txt")
if not family in d_count_pr[lig_ref].keys () :
d_count_pr[lig_ref][family] = 0
d_count_pr[lig_ref][family] = d_count_pr[lig_ref][family] + 1
except: pass
# BS -> query
for file_BS in l_file_BS :
# for not reference BS
if not search ("BS_REF", file_BS) :
lig_querie = file_BS.split ("_")[1]
prot_querie = file_BS.split ("_")[2][0:4]
print prot_querie, lig_querie, "*******"
# find ligand reference
# lig ref define in previous step
d_count_pr[lig_ref]["pr queries"].append (prot_querie)
d_count_pr[lig_ref]["lig queries"].append (lig_querie)
# write and plot #
##################
pr_result = pathManage.generatePath(pr_final_folder + "counting/")
for ligand_sub in d_count.keys () :
p_filout = pr_result + ligand_sub
filout = open (p_filout, "w")
filout.write ("\t".join(d_count[ligand_sub].keys ()) + "\n")
l_value = [str(x) for x in d_count[ligand_sub].values ()]
filout.write ("\t".join(l_value) + "\n")
filout.close ()
runOtherSoft.piePlot(p_filout)
filout_lig = open (pr_result + "count_ligand", "w")
filout_lig.write ("Ligand ID\tNumber of occurences in the dataset\tNumber of different clusters\tList of clusters\tList of protein families\n")
for lig in d_lig.keys () :
if d_lig[lig] > 1 :
filout_lig.write (str (lig) + "\t" + str (d_lig[lig]["count"]) + "\t" + str(len (list (set(d_lig[lig]["group"])))) + "\t" + " ".join (d_lig[lig]["group"]) + "\t" + " ".join (d_lig[lig]["family"]) + "\n")
filout_lig.close ()
filout_LSR_lig = open (pr_result + "CountByLigandRef", "w")
for lig_ref in d_by_ref.keys () :
filout_LSR_lig.write ("====" + str (lig_ref) + "====\n")
for sub_ref in d_by_ref[lig_ref].keys () :
filout_LSR_lig.write (str (sub_ref) + ": " + str (d_by_ref[lig_ref][sub_ref]) + "\n")
filout_LSR_lig.close ()
filout_pr_count = open (pr_result + "count_pr", "w")
for lig in d_count_pr.keys () :
filout_pr_count.write ("====" + str (lig) + "====\n")
filout_pr_count.write ("nb ref pr: " + str (len (d_count_pr[lig]["pr ref"])) + "\n")
filout_pr_count.write ("nb querie pr: " + str (len (d_count_pr[lig]["pr queries"])) + "\n")
filout_pr_count.write ("nb ligand queries: " + str (len (d_count_pr[lig]["lig queries"])) + "\n")
for family in d_count_pr[lig].keys () :
if family != "pr ref" and family != "pr queries" and family != "lig queries" :
filout_pr_count.write ("Ref " + str (family) + ": " + str (d_count_pr[lig][family]) + "\n")
filout_pr_count.close ()
runOtherSoft.barplot(pr_result + "count_ligand")
def qualityExtraction (l_ligand, name_folder, p_list_ligand, thresold_sheap) :
pr_result = pathManage.result("final_" + name_folder)
filout = open(pr_result + "quality_extraction.txt", "w")
# number PDB by ligand, without filter
filout.write ("Number PDB by ligand:\n")
d_dataset = tool.parseLigandPDBList(p_list_ligand)
for ligand in l_ligand :
filout.write (str (ligand) + ": " + str (len (d_dataset[ligand])) + "\n")
# number references
filout.write ("\n*************\n\nNumber references by ligands:\n")
for ligand in l_ligand :
pr_result_ligand = pathManage.result(ligand)
nb_ref = -2
l_file = listdir(pr_result_ligand)
for f in l_file :
if path.isdir (pr_result_ligand + "/" + f) :
nb_ref = nb_ref + 1
filout.write (ligand + ": " + str (nb_ref) + "\n")
# number of query by ref in means and max and min (after blast)
filout.write ("\n*************\n\nNumber means queries by references:\n")
p_family_all = pathManage.result() + "reference_family_all.txt"
filout_family_all = open (p_family_all, "w")
d_family_all = {}
for ligand in l_ligand :
d_nb_query = {}
d_family = {}
p_filout_family = pathManage.result() + "reference_family_" + ligand + ".txt"
p_filout_family_count = pathManage.result () + "count_family_" + ligand + ".txt"
filout_family = open (p_filout_family, "w")
filout_family_count = open (p_filout_family_count, "w")
pr_result_ligand = pathManage.result(ligand)
nb_ref = 0
l_file = listdir(pr_result_ligand)
for f in l_file :
if path.isdir (pr_result_ligand + "/" + f) and len (f) == 4:
# count by family
family_ref = analysis.findFamily(f, pathManage.findFamilyFile (ligand))
filout_family.write ("\t".join (family_ref) + "\n")
if not family_ref[-1] in d_family.keys () :
d_family[family_ref[-1]] = 0
d_family[family_ref[-1]] = d_family[family_ref[-1]] + 1
# file all
if not family_ref[-1] in d_family_all.keys () :
d_family_all[family_ref[-1]] = 0
d_family_all[family_ref[-1]] = d_family_all[family_ref[-1]] + 1
# count number of references
nb_ref = nb_ref + 1
d_nb_query[f] = 0
l_file_queries = listdir(pr_result_ligand + "/" + f + "/")
for file_query in l_file_queries :
if search ("CX",file_query) :
d_nb_query[f] = d_nb_query[f] + 1
filout.write (ligand + ": " + str(np.sum(d_nb_query.values ())) + "\n")
filout.write (ligand + ": " + str(np.mean(d_nb_query.values ())) + "+/-" + str(np.std (d_nb_query.values ())) + "\n")
filout.write ("MAX " + str (ligand) + ": " + str (max (d_nb_query.values ())) + " " + str (d_nb_query.keys ()[d_nb_query.values ().index (max (d_nb_query.values ()))]) +"\n")
# family
filout_family_count.write ("\t".join(d_family.keys ()) + "\n")
l_values = [str(x) for x in d_family.values ()]
filout_family_count.write ("\t".join(l_values) + "\n")
filout_family.close ()
filout_family_count.close ()
runOtherSoft.piePlot(p_filout_family_count)
# all family
filout_family_all.write ("\t".join(d_family_all.keys ()) + "\n")
l_values = [str(x) for x in d_family_all.values ()]
filout_family_all.write ("\t".join(l_values) + "\n")
filout_family_all.close ()
runOtherSoft.piePlot(p_family_all)
# number subref by ligand
filout.write ("\n*************\n\nNumber of subref considered:\n")
for ligand in l_ligand :
d_nb_sub = {}
d_nb_sub_sheap = {}
pr_result_ligand = pathManage.result(ligand)
l_ref = listdir(pr_result_ligand)
for ref in l_ref :
if path.isdir (pr_result_ligand + "/" + ref) and len (ref) == 4:
l_file_queries = listdir(pr_result_ligand + "/" + ref + "/")
for file_query in l_file_queries :
if search ("substituent",file_query) and search (".pdb",file_query):
atom_substituate = file_query.split ("_")[-2]
try : value_sheap = float(file_query.split ("_")[-1][:-4])
except : continue
if not atom_substituate in d_nb_sub.keys () :
d_nb_sub[atom_substituate] = 0
d_nb_sub[atom_substituate] = d_nb_sub[atom_substituate] + 1
if value_sheap > thresold_sheap :
if not atom_substituate in d_nb_sub_sheap :
d_nb_sub_sheap[atom_substituate] = 0
d_nb_sub_sheap[atom_substituate] = d_nb_sub_sheap[atom_substituate] + 1
filout.write ("\n" + ligand + "\n")
for atom_substituate in d_nb_sub.keys () :
filout.write (atom_substituate + ": " + str (d_nb_sub[atom_substituate]) + "\n")
try : filout.write (atom_substituate + " ShaEP: " + str (d_nb_sub_sheap[atom_substituate]) + "\n")
except : filout.write (atom_substituate + " ShaEP: 0\n")
filout.close()