def formatOPERAForToolChem(self):
# to change
c_chem = CompDesc.CompDesc("", "")
l_desc2D = c_chem.getLdesc("2D")
pr_desc = self.pr_out + "DESC/"
l_file_chem = listdir(self.pr_forToolChem)
for file_chem in l_file_chem:
if file_chem != "chemicals_listNew.csv": ##############################################################
continue ##########################################################################################
p_filout = self.pr_forToolChem + file_chem[:-4] + "_desc2D.csv"
filout = open(p_filout, "w")
filout.write("inchikey\t" + "\t".join(l_desc2D))
l_chem = toolbox.loadMatrixToList(self.pr_forToolChem + file_chem)
i = 0
imax = len(l_chem)
while i < imax:
d_chem = l_chem[i]
c_chem = CompDesc.CompDesc(d_chem["smiles_origin"], pr_desc)
c_chem.prepChem()
if c_chem.err == 0:
c_chem.computeAll2D()
if c_chem.err == 0:
filout.write("%s\t%s\n" % (c_chem.inchikey, "\t".join(
[str(c_chem.all2D[desc]) for desc in l_desc2D])))
i = i + 1
filout.close()
def test_generate3D(self):
cChem = CompDesc.CompDesc("N=C(O)[C@@H](N)CS", "./tests/")
cChem.prepChem()
cChem.set3DChemical()
self.assertEqual(cChem.err, 0)
rmtree("./tests/MOLCLEAN")
rmtree("./tests/SDF3D")
def computeMissingPNG(self):
pr_organisedPNG = pathFolder.createFolder(self.pr_out +
"PNGorganized/")
pr_temp = pathFolder.createFolder(self.pr_out + "temp_PNG/")
# load from table
l_chem_chemicalsDB = self.cDB.execCMD(
"SELECT DISTINCT smiles_clean, inchikey FROM chemicals WHERE inchikey is not null AND smiles_clean is not NULL"
) # extract dsstoxID when inchikey is not null
shuffle(l_chem_chemicalsDB)
i = 0
imax = len(l_chem_chemicalsDB)
compute = 0
while i < imax:
if i % 100 == 0:
print(i, compute)
smiles_clean = l_chem_chemicalsDB[i][0]
inchikey = l_chem_chemicalsDB[i][1]
p_png = pr_organisedPNG + inchikey[:2] + "/" + inchikey[
2:4] + "/" + inchikey + ".png"
if not path.exists(p_png):
cChem = CompDesc.CompDesc(smiles_clean, pr_temp)
cChem.inchikey = inchikey
cChem.smi = smiles_clean
cChem.computePNG(bg="none")
p_png_temp = pr_temp + "PNG/" + inchikey + ".png"
if path.exists(p_png_temp) and path.getsize(p_png_temp) > 0:
pathFolder.createFolder(pr_organisedPNG + inchikey[:2] +
"/" + inchikey[2:4] + "/")
copyfile(p_png_temp, p_png)
compute = compute + 1
i = i + 1
def formatPrepChemForToolChem(self):
pr_desc = self.pr_out + "DESC/"
l_file_chem = listdir(self.pr_forToolChem)
for file_chem in l_file_chem:
if file_chem != "chemicals_listNew.csv": ##############################################################
continue ##########################################################################################
p_filout = self.pr_forToolChem + file_chem[:-4] + "_chemPrep.csv"
filout = open(p_filout, "w")
filout.write(
"dsstox_id\tsmiles_origin\tsmiles_cleaned\tinchikey\tdrugbank_id\tcasn\tname\n"
)
l_chem = toolbox.loadMatrixToList(self.pr_forToolChem + file_chem)
i = 0
imax = len(l_chem)
while i < imax:
d_chem = l_chem[i]
c_chem = CompDesc.CompDesc(d_chem["smiles_origin"], pr_desc)
c_chem.prepChem()
if c_chem.err == 0:
c_chem.computeAll2D()
if c_chem.err == 0:
filout.write(
"%s\t%s\t%s\t%s\t%s\t%s\t%s\n" %
(d_chem["dsstox_id"], d_chem["smiles_origin"],
c_chem.smi, c_chem.inchikey, "NA", d_chem["casn"],
d_chem["name"]))
i = i + 1
filout.close()
def updateSMILES(self, name_table="chemicals"):
"""Function use to update the chemical table => check if smiles origin change"""
d_dsstox_SMILES = toolbox.loadMatrixToDict(self.p_chem_SMILES, sep=",")
d_dsstox_name = toolbox.loadMatrixToDict(self.p_chem_name)
self.pr_desc = pathFolder.createFolder(self.pr_out + "DESC/")
#extract chemical without DTXSID
# see if chem included
cmd_SQL = "SELECT id, dsstox_id, smiles_origin, inchikey, smiles_clean FROM %s " % (
name_table)
l_chem_DB = self.cDB.execCMD(cmd_SQL)
d_chem_DB = {}
for chem_DB in l_chem_DB:
d_chem_DB[chem_DB[1]] = [
chem_DB[0], chem_DB[2], chem_DB[3], chem_DB[4]
]
i = 0
for chem in d_dsstox_SMILES.keys():
dsstox_id = d_dsstox_SMILES[chem]["dsstox_substance_id"]
inchkey = d_dsstox_SMILES[chem]["InChI Key_QSARr"]
smiles = d_dsstox_SMILES[chem]["Original_SMILES"]
smiles_cleaned = d_dsstox_SMILES[chem]["Canonical_QSARr"]
try:
smiles_indb = d_chem_DB[dsstox_id][
1] # case of chemical is not in the DB
except:
continue
inchkey_db = d_chem_DB[dsstox_id][2]
smiles_cleaned_db = d_chem_DB[dsstox_id][3]
smiles_db = d_chem_DB[dsstox_id][1]
if smiles != smiles_db:
# recompute cleaned SMILES
c_chem = CompDesc.CompDesc(smiles, self.pr_desc)
c_chem.prepChem()
if c_chem.err == 0:
c_chem.generateInchiKey()
else:
c_chem.smi = None
if c_chem.err == 0:
inchikey = c_chem.inchikey
else:
inchikey = None
if d_chem_DB[dsstox_id][2] != inchikey:
cmd_sql = "UPDATE %s SET smiles_origin = '%s', smiles_clean = '%s', inchikey='%s' WHERE id='%s';" % (
name_table, smiles, c_chem.smi, inchikey,
d_chem_DB[dsstox_id][0])
else:
continue #cmd_sql = "UPDATE %s SET smiles_origin = '%s' WHERE id='%s';"%(name_table, smiles, d_chem_DB[dsstox_id][0])
#print(smiles_cleaned,smiles_indb, dsstox_id)
print(i)
i = i + 1
self.cDB.updateTable(cmd_sql)
return
def test_FP(self):
cChem = CompDesc.CompDesc("N=C(O)[C@@H](N)CS", "./tests/")
cChem.prepChem()
cChem.computeFP("All")
# test comparison
cCchem2 = CompDesc.CompDesc("CCCO", "./tests/")
cCchem2.prepChem()
cCchem2.computeFP("All")
l_dist = ["Tanimoto", "Dice", "Cosine", "Sokal", "Russel", "RogotGoldberg", "Kulczynski", "McConnaughey", "Asymmetric", "BraunBlanquet", "AllBit"]
l_fp = ['Mol', 'pairs', 'MACCS', 'Torsion', 'Morgan']
for fp in l_fp:
for dist in l_dist:
print(fp, dist, cChem.computeSimilarityFP(cCchem2, fp, dist))
# test all combination
self.assertEqual(cChem.err, 0)
def test_computeOPERAServer(self):
cChem = CompDesc.CompDesc("N=C(O)[C@@H](N)CS", "./tests/")
cChem.prepChem()
cChem.computePADEL2DFPandCDK()
cChem.computeOperaDesc()
self.assertEqual(cChem.err, 0)
system("rm -rf ./tests/PADEL*")
system("rm -rf ./tests/cdk_desc*")
rmtree("./tests/OPERA")
def test_computeOPERA(self):
cChem = CompDesc.CompDesc("N=C(O)[C@@H](N)CS", "./tests/")
cChem.prepChem()
cChem.computeOPERAFromChem(update=1)
self.assertEqual(cChem.err, 0)
system("rm -rf ./tests/PADEL*")
system("rm -rf ./tests/CDK*")
system("rm -rf ./tests/cdk_desc*")
rmtree("./tests/OPERA")
def test_compute3Ddesc(self):
cChem = CompDesc.CompDesc("CC(C)C[C@H](NC(=O)[C@H](Cc1ccccc1)NC(=O)[C@H](CCC(=O)O)NC(=O)[C@@H](N)CCCCN)C(=O)N[C@@H](Cc1cnc[nH]1)C(=O)N1CCC[C@H]1C(=O)N[C@@H](CO)C(=O)N[C@@H](CCCCN)C(=O)N[C@H](C(=O)N[C@@H](CC(=O)O)C(=O)N[C@@H](CC(C)C)C(=O)N1CCC[C@H]1C(=O)N[C@@H](CCCNC(=N)N)C(=O)O)C(C)C", "./tests/")
cChem.prepChem()
cChem.set3DChemical()
cChem.computeAll3D()
self.assertEqual(cChem.err, 0)
rmtree("./tests/MOLCLEAN")
rmtree("./tests/SDF3D")
rmtree("./tests/3D")
def test_knimeConvert(self):
cChem = CompDesc.CompDesc("N=C(O)[C@@H](N)CS", "./tests/")
cChem.prepChem()
cChem.computeAll2D()
cChem.convertDesc2DtoKnimeDesc()
err = 0
try: test = cChem.all2D["AMW"]
except:err = 1
self.assertEqual(err, 0)
def computeDescNewChem(self):
if not "l_chem_toadd" in self.__dict__:
self.extractOnlyNewChem("chemicals", "dsstox_id")
self.pr_desc = pathFolder.createFolder(self.pr_out + "DESC/")
d_dsstox_SMILES = toolbox.loadMatrixToDict(self.p_chem_SMILES, sep=",")
l_chem_add = self.l_chem_toadd
shuffle(l_chem_add)
i = 0
imax = len(self.l_chem_toadd)
print(imax)
while i < imax:
if i % 1000 == 0:
print(i)
chem = l_chem_add[i]
try:
smiles = d_dsstox_SMILES[chem]["Original_SMILES"]
except:
print(i, ": ERROR in SMILES - ", chem)
i = i + 1
continue
cChem = CompDesc.CompDesc(smiles, self.pr_desc)
cChem.prepChem()
if cChem.err == 0:
cChem.generateInchiKey()
if cChem.err == 1:
print("Error inch: %s" % (l_chem_add[i]))
i = i + 1
continue
# 2D desc
cChem.computeAll2D()
cChem.writeMatrix("2D")
#3D desc
cChem.set3DChemical()
if cChem.err == 0:
cChem.computeAll3D()
if cChem.err == 1:
print("Error 3D desc: %s -- %s" % (l_chem_add[i], i))
else:
cChem.writeMatrix("3D")
else:
print("Error 3D generation: %s -- %s" % (l_chem_add[i], i))
else:
print("Error prep: %s -- %s" % (l_chem_add[i], i))
i = i + 1
def test_PrepChem(self):
cChem = CompDesc.CompDesc("O=S(=O)(O)c1ccc([Hg])cc1", "./tests/")
cChem.prepChem()
smi = cChem.smi
self.assertEqual(smi, "O=S(=O)(O)c1ccc([Hg])cc1")
def test_compute1D2Ddesc(self):
cChem = CompDesc.CompDesc("N=C(O)[C@@H](N)CS", "./tests/")
cChem.prepChem()
cChem.computeAll2D()
self.assertEqual(cChem.err, 0)
def pushNewChemInDB(self, name_table="chemicals"):
if not "l_chem_toadd" in self.__dict__:
self.extractOnlyNewChem(name_table, field_comparison)
d_dsstox_name = toolbox.loadMatrixToDict(self.p_chem_name)
d_dsstox_SMILES = toolbox.loadMatrixToDict(self.p_chem_SMILES, sep=",")
self.pr_desc = pathFolder.createFolder(self.pr_out + "DESC/")
id_chem = self.cDB.execCMD("SELECT MAX(id) FROM %s" %
(name_table))[0][0]
i = 0
imax = len(self.l_chem_toadd)
#imax = 100
print(imax)
cmd_all = ""
while i < imax:
if i % 1000 == 0:
print(i)
chem = self.l_chem_toadd[i]
# case of 50K chemicals without SMILES
try:
smiles = d_dsstox_SMILES[chem]["Original_SMILES"]
except:
smiles = ""
name = d_dsstox_name[chem]["preferred_name"]
casrn = d_dsstox_name[chem]["casrn"]
cChem = CompDesc.CompDesc(smiles, self.pr_desc)
if smiles != "":
cChem.prepChem()
else:
cChem.err = 1
if cChem.err == 0:
smi_cleaned = cChem.smi
if cChem.err == 1:
inch = ""
else:
inch = cChem.generateInchiKey()
else:
smi_cleaned = ""
inch = ""
id_chem = id_chem + 1
if smiles == "":
cmd_sql = "INSERT INTO %s (id, dsstox_id, casn, name) VALUES (%s, '%s', '%s', '%s');\n" % (
name_table, id_chem, chem, casrn, name.replace("'", "''"))
elif smi_cleaned == "":
cmd_sql = "INSERT INTO %s (id, smiles_origin, dsstox_id, casn, name) VALUES (%s, '%s', '%s', '%s', '%s');\n" % (
name_table, id_chem, smiles, chem, casrn,
name.replace("'", "''"))
elif inch == "":
cmd_sql = "INSERT INTO %s (id, smiles_origin, smiles_clean, dsstox_id, casn, name) VALUES (%s, '%s', '%s', '%s', '%s', '%s');\n" % (
name_table, id_chem, smiles, smi_cleaned, chem, casrn,
name.replace("'", "''"))
else:
cmd_sql = "INSERT INTO %s (id, smiles_origin, smiles_clean, inchikey, dsstox_id, casn, name) VALUES (%s, '%s', '%s', '%s', '%s', '%s', '%s');\n" % (
name_table, id_chem, smiles, smi_cleaned, inch, chem,
casrn, name.replace("'", "''"))
i = i + 1
cmd_all = cmd_all + cmd_sql
self.cDB.runCMDaddElement(cmd_all)