def FragmentUnsantize(suppl1):
try:
newmol = Chem.FragmentOnBRICSBonds(suppl1)
#print('here')
mfl = Chem.GetMolFrags(newmol, asMols=True, sanitizeFrags=False)
print('Good False')
return mfl
except:
print('Not good for false')
raise RDKitError(2)
def FragmentSanitize(tempSDFPath):
try:
suppl2 = Chem.SDMolSupplier(tempSDFPath, sanitize=True)
newmol2 = Chem.FragmentOnBRICSBonds(suppl2[0])
mfl = Chem.GetMolFrags(newmol2, asMols=True, sanitizeFrags=False)
#print('Good True')
return mfl
except:
#print('Not good for true')
raise RDKitError(1)
def BreakBRICSBonds(mol, bonds=None, sanitize=True, silent=True):
""" breaks the BRICS bonds in a molecule and returns the results
>>> from rdkit import Chem
>>> m = Chem.MolFromSmiles('CCCOCC')
>>> m2=BreakBRICSBonds(m)
>>> Chem.MolToSmiles(m2,True)
'[3*]O[3*].[4*]CC.[4*]CCC'
a more complicated case:
>>> m = Chem.MolFromSmiles('CCCOCCC(=O)c1ccccc1')
>>> m2=BreakBRICSBonds(m)
>>> Chem.MolToSmiles(m2,True)
'[16*]c1ccccc1.[3*]O[3*].[4*]CCC.[4*]CCC([6*])=O'
can also specify a limited set of bonds to work with:
>>> m = Chem.MolFromSmiles('CCCOCC')
>>> m2 = BreakBRICSBonds(m,[((3, 2), ('3', '4'))])
>>> Chem.MolToSmiles(m2,True)
'[3*]OCC.[4*]CCC'
this can be used as an alternate approach for doing a BRICS decomposition by
following BreakBRICSBonds with a call to Chem.GetMolFrags:
>>> m = Chem.MolFromSmiles('CCCOCC')
>>> m2=BreakBRICSBonds(m)
>>> frags = Chem.GetMolFrags(m2,asMols=True)
>>> [Chem.MolToSmiles(x,True) for x in frags]
['[4*]CCC', '[3*]O[3*]', '[4*]CC']
"""
if not bonds:
#bonds = FindBRICSBonds(mol)
res = Chem.FragmentOnBRICSBonds(mol)
if sanitize:
Chem.SanitizeMol(res)
return res
eMol = Chem.EditableMol(mol)
nAts = mol.GetNumAtoms()
dummyPositions = []
for indices, dummyTypes in bonds:
ia, ib = indices
obond = mol.GetBondBetweenAtoms(ia, ib)
bondType = obond.GetBondType()
eMol.RemoveBond(ia, ib)
da, db = dummyTypes
atoma = Chem.Atom(0)
atoma.SetIsotope(int(da))
atoma.SetNoImplicit(True)
idxa = nAts
nAts += 1
eMol.AddAtom(atoma)
eMol.AddBond(ia, idxa, bondType)
atomb = Chem.Atom(0)
atomb.SetIsotope(int(db))
atomb.SetNoImplicit(True)
idxb = nAts
nAts += 1
eMol.AddAtom(atomb)
eMol.AddBond(ib, idxb, bondType)
if mol.GetNumConformers():
dummyPositions.append((idxa, ib))
dummyPositions.append((idxb, ia))
res = eMol.GetMol()
if sanitize:
Chem.SanitizeMol(res)
if mol.GetNumConformers():
for conf in mol.GetConformers():
resConf = res.GetConformer(conf.GetId())
for ia, pa in dummyPositions:
resConf.SetAtomPosition(ia, conf.GetAtomPosition(pa))
return res
def ChopWithRDKit(outputDir, inputPath):
#read input from terminal and get file name
lig = os.path.basename(inputPath) #file name, no path
#output folder
output = outputDir + 'output-chop/'
outputFolderPath_log = outputDir + 'output-log/'
outputFolderPath_sdf = outputDir + 'output-sdf/'
outputFolderPath_chop_comb = outputDir + 'output-chop-comb/'
suppl = Chem.MolFromMol2File(inputPath, sanitize=False)
newmol = Chem.FragmentOnBRICSBonds(suppl)
mfl = Chem.GetMolFrags(newmol, asMols=True, sanitizeFrags=False)
tempSDFPath = outputDir + 'output-sdf/' + lig + '.sdf'
w = Chem.SDWriter(tempSDFPath)
w.SetKekulize(False)
w.write(suppl)
w.close()
#generate fragments with rdkit
fileList = []
f = 0
l = 0
r = 0
for m in mfl:
carbonC = 0
nitrogC = 0
oxygenC = 0
rmAtomCount = 0
for i in range(m.GetNumAtoms()):
#record dummy atom and hydrogen number
if m.GetAtomWithIdx(i).GetSymbol() == '*':
rmAtomCount = rmAtomCount + 1
if m.GetAtomWithIdx(i).GetSymbol() == 'H':
rmAtomCount = rmAtomCount + 1
if m.GetAtomWithIdx(i).GetSymbol() == 'C':
carbonC = carbonC + 1
if m.GetAtomWithIdx(i).GetSymbol() == 'N':
nitrogC = nitrogC + 1
if m.GetAtomWithIdx(i).GetSymbol() == 'O':
oxygenC = oxygenC + 1
#create file
totalAtomNum = m.GetNumAtoms() - rmAtomCount
if m.GetNumAtoms() - rmAtomCount >= 4:
tempFileName = output + 'b-' + lig + '-' + str(r).zfill(3) + '.sdf'
r = r + 1
if m.GetNumAtoms() - rmAtomCount < 4:
tempFileName = output + 'l-' + lig + '-' + str(l).zfill(3) + '.sdf'
l = l + 1
w = Chem.SDWriter(tempFileName)
w.SetKekulize(False)
f = f + 1
w.write(m)
w.close()
fileList.append(tempFileName)
#create file list with atom numbers
with open(outputFolderPath_log + 'ListAll', 'at') as outlist:
outlist.write(tempFileName + ' T ' + str(totalAtomNum) + ' C ' +
str(carbonC) + ' N ' + str(nitrogC) + ' O ' +
str(oxygenC) + ' \n')
#with open(outputFolderPath_log+'Process.log','at') as outf:
# outf.write('Files are created.\n')
#read atom coordinates and atom type from mol2 file
mol2AllList = []
with open(inputPath, 'r') as inf:
mol2AllList = inf.readlines()
mol2AtomInfo = []
molHead = mol2AllList.index('@<TRIPOS>ATOM\n')
molEnd = mol2AllList.index('@<TRIPOS>BOND\n')
mol2AtomInfo = mol2AllList[molHead + 1:molEnd]
mol2X = []
mol2Y = []
mol2Z = []
mol2A = []
for i in range(len(mol2AtomInfo)):
mol2Line = mol2AtomInfo[i].split()
mol2X.append(float(mol2Line[2]))
mol2Y.append(float(mol2Line[3]))
mol2Z.append(float(mol2Line[4]))
mol2A.append(mol2Line[5])
for filePath in fileList:
fileName = os.path.basename(filePath)
if len(fileName) > 0:
#processing brick fragments
if fileName[0] == 'b':
brickInfoList = []
with open(filePath, 'r') as inf:
brickInfoList = inf.readlines()
#print(brickInfoList)
brickMolEndList = [
i for i, x in enumerate(brickInfoList) if x == '$$$$\n'
]
#print(brickInfoList[:brickMolEndList[0]])
fileHead = list(filter(lambda x: 'V2000' in x, brickInfoList))
fileHeadLineNum = brickInfoList.index(fileHead[0])
#print(fileHeadLineNum)
fileHeadList = fileHead[0].split()
atomNum = int(fileHead[0][0:3])
bondNum = int(fileHead[0][3:6])
atomList = brickInfoList[fileHeadLineNum + 1:fileHeadLineNum +
atomNum + 1]
bondList = brickInfoList[fileHeadLineNum + atomNum +
1:fileHeadLineNum + atomNum +
bondNum + 1]
#Search for atom type
atomTypeList = []
dummyAtomList = []
dummyAtomLineList = []
hydrAtomList = []
hydrAtomLineList = []
for atomLine in atomList:
atomLineInfoList = atomLine.split()
#atom in brick.sdf, xyz coordinates of one line
atomX = float(atomLineInfoList[0])
atomY = float(atomLineInfoList[1])
atomZ = float(atomLineInfoList[2])
#calculate norm
normList = []
for i in range(len(mol2AtomInfo)):
norm = (atomX - mol2X[i]) * (atomX - mol2X[i]) + (
atomY - mol2Y[i]) * (atomY - mol2Y[i]) + (
atomZ - mol2Z[i]) * (atomZ - mol2Z[i])
normList.append(norm)
minInd = normList.index(min(normList))
atomTypeList.append(mol2A[minInd] + '\n')
#dummy atom List
if atomLineInfoList[3] == "R":
dummyAtomList.append(atomList.index(atomLine))
dummyAtomLineList.append(atomLine)
#hydrogen atom list
if atomLineInfoList[3] == "H":
hydrAtomList.append(atomList.index(atomLine))
hydrAtomLineList.append(atomLine)
newBrickInfoList = brickInfoList[:brickMolEndList[0]]
#Branch, eligible to connect
bondInfoList = []
for bondLine in bondList:
#bondLineInfoList=bondLine.split()
bondLineInfoList = [bondLine[0:3], bondLine[3:6]
] + bondLine[6:].split()
bondInfoList.append(
[int(bondLineInfoList[0]),
int(bondLineInfoList[1])])
dummyConnection = [
] #dummyConnection is a list of connections of the original file, eg. ['8 14 1 0\n',''], which will be used to remove not using connections in the last step
allConnection = [
] #all connection is a list of connection pairs, eg. [[8,14],[6,15]], which will be used to generate appendix II
for dummyIdx in dummyAtomList:
connectionList = list(
filter(lambda x: dummyIdx + 1 in x, bondInfoList))
for tempCon in connectionList:
conIndex = bondInfoList.index(tempCon)
dummyConnection.append(bondList[conIndex])
#remove the case both dummy atoms are in the bond
rmBond = []
for connect in connectionList:
if connect[0] - 1 in dummyAtomList:
if connect[1] - 1 in dummyAtomList:
rmBond.append(connect)
for tempBond in rmBond:
connectionList.remove(tempBond)
allConnection = allConnection + connectionList
tempDummyCon = []
for dummyCon in dummyConnection:
if dummyCon not in tempDummyCon:
tempDummyCon.append(dummyCon)
dummyConnection = tempDummyCon
branchCon = []
for connect in allConnection:
if connect[0] - 1 in dummyAtomList:
branchCon.append(
str(connect[1]) + ' ' +
atomTypeList[connect[0] - 1])
if connect[1] - 1 in dummyAtomList:
branchCon.append(
str(connect[0]) + ' ' +
atomTypeList[connect[1] - 1])
#sort branch by atom index
branchConAtomList = []
branchConAtomListBefore = []
branchConAtomIndexList = []
newBranchCon = [] #newBranchCon is the list of appendix II
for branchLine in branchCon:
branchLineList = branchLine.split()
branchConAtomListBefore.append(branchLineList[0])
branchConAtomList = sorted(branchConAtomListBefore)
branchConAtomIndexList = sorted(
range(len(branchConAtomListBefore)),
key=lambda k: branchConAtomListBefore[k])
for ind in range(len(branchConAtomList)):
newBranchCon.append(branchCon[branchConAtomIndexList[ind]])
#hydrogen
hydrConnection = []
for hydrIdx in hydrAtomList:
connectionList = filter(lambda x: hydrIdx + 1 in x,
bondInfoList)
for tempCon in connectionList:
conIndex = bondInfoList.index(tempCon)
hydrConnection.append(bondList[conIndex])
#edit head line
newAtomNum = atomNum - len(dummyAtomLineList) - len(
hydrAtomLineList)
newBondNum = bondNum - len(dummyConnection) - len(
hydrConnection)
newHead = str(newAtomNum).rjust(3) + str(newBondNum).rjust(
3) + fileHead[0][6:]
newBrickInfoList[fileHeadLineNum] = newHead
newBrickInfoList[0] = fileName + '\n'
#edit output list
#edit appendix I - ATOM TYPES
newBrickInfoList.append('\n')
newBrickInfoList.append('> <ATOMTYPES> \n')
newBrickInfoList = newBrickInfoList + atomTypeList[:newAtomNum]
#edit appendix II - BRANCH ATOM NUMBER AND ELIGIBLE ATMTYPE TO CONNECT
newBrickInfoList.append('\n')
newBrickInfoList.append(
'> <BRANCH @atom-number eligible-atmtype-to-connect> \n')
newBrickInfoList = newBrickInfoList + newBranchCon
newBrickInfoList.append('\n')
newBrickInfoList.append('$$$$\n')
#remove dummy atoms
for dummyLine in dummyAtomLineList:
newBrickInfoList.remove(dummyLine)
#remove dummy bonds
for dummyCon in dummyConnection:
newBrickInfoList.remove(dummyCon)
#remove hydrogen atoms
for hydrLine in hydrAtomLineList:
newBrickInfoList.remove(hydrLine)
#remove hydrogen bonds
for hydrCon in hydrConnection:
newBrickInfoList.remove(hydrCon)
#remove M ISO line
fileMISO = list(
filter(lambda x: 'M ISO' in x, newBrickInfoList))
#print(fileMISO)
if len(fileMISO) > 0:
for ISO in fileMISO:
newBrickInfoList.remove(ISO)
#remove M CHG line
fileMCHG = list(
filter(lambda x: 'M CHG' in x, newBrickInfoList))
if len(fileMCHG) > 0:
for CHG in fileMCHG:
newBrickInfoList.remove(CHG)
#write brick info to file
with open(filePath, 'w') as outf:
outf.writelines(newBrickInfoList)
#Processing linker fragments
if fileName[0] == 'l':
linkerInfoList = []
with open(filePath, 'r+') as inf:
linkerInfoList = inf.readlines()
#find the end of molcules
linkerMolEndList = [
i for i, x in enumerate(linkerInfoList) if x == '$$$$\n'
]
#find the start of molecules
fileHead = list(filter(lambda x: 'V2000' in x, linkerInfoList))
#indicate the line num of the head line
fileHeadLineNum = linkerInfoList.index(fileHead[0])
#separate atom num and bond num, then separate atom and bond info
fileHeadList = fileHead[0].split()
atomNum = int(fileHead[0][0:3])
bondNum = int(fileHead[0][3:6])
atomList = linkerInfoList[fileHeadLineNum + 1:fileHeadLineNum +
atomNum + 1]
bondList = linkerInfoList[fileHeadLineNum + atomNum +
1:fileHeadLineNum + atomNum +
bondNum + 1]
#Search for atom type
atomTypeList = []
dummyAtomList = []
dummyAtomLineList = []
hydrAtomList = []
hydrAtomLineList = []
for atomLine in atomList:
atomLineInfoList = atomLine.split()
#atom in brick.sdf, xyz coordinates of one line
atomX = float(atomLineInfoList[0])
atomY = float(atomLineInfoList[1])
atomZ = float(atomLineInfoList[2])
#calculate norm
normList = []
for i in range(len(mol2AtomInfo)):
norm = (atomX - mol2X[i]) * (atomX - mol2X[i]) + (
atomY - mol2Y[i]) * (atomY - mol2Y[i]) + (
atomZ - mol2Z[i]) * (atomZ - mol2Z[i])
normList.append(norm)
minInd = normList.index(min(normList))
atomTypeList.append(mol2A[minInd] + '\n')
#dummy atom List
if atomLineInfoList[3] == "R":
dummyAtomList.append(atomList.index(atomLine))
dummyAtomLineList.append(atomLine)
#hydrogen atom list
if atomLineInfoList[3] == "H":
hydrAtomList.append(atomList.index(atomLine))
hydrAtomLineList.append(atomLine)
newLinkerInfoList = linkerInfoList[:linkerMolEndList[0]]
#Branch, eligible to connect
bondInfoList = []
for bondLine in bondList:
#bondLineInfoList=bondLine.split()
bondLineInfoList = [bondLine[0:3], bondLine[3:6]
] + bondLine[6:].split()
bondInfoList.append(
[int(bondLineInfoList[0]),
int(bondLineInfoList[1])])
dummyConnection = []
allConnection = []
for dummyIdx in dummyAtomList:
connectionList = list(
filter(lambda x: dummyIdx + 1 in x, bondInfoList))
for tempCon in connectionList:
conIndex = bondInfoList.index(tempCon)
dummyConnection.append(bondList[conIndex])
#remove the case both dummy atom are in the bond
rmBond = []
for connect in connectionList:
if connect[0] - 1 in dummyAtomList:
if connect[1] - 1 in dummyAtomList:
rmBond.append(connect)
for tempBond in rmBond:
connectionList.remove(tempBond)
allConnection = allConnection + connectionList
#remove connection duplicates
tempDummyCon = []
for dummyCon in dummyConnection:
if dummyCon not in tempDummyCon:
tempDummyCon.append(dummyCon)
dummyConnection = tempDummyCon
contactCount = []
for connect in allConnection:
if connect[0] - 1 in dummyAtomList:
contactCount.append(connect[1])
if connect[1] - 1 in dummyAtomList:
contactCount.append(connect[0])
#hydrogen
hydrConnection = []
for hydrIdx in hydrAtomList:
connectionList = list(
filter(lambda x: hydrIdx + 1 in x, bondInfoList))
for tempCon in connectionList:
conIndex = bondInfoList.index(tempCon)
hydrConnection.append(bondList[conIndex])
#edit output list
#edit head line
newAtomNum = atomNum - len(dummyAtomLineList) - len(
hydrAtomLineList)
newBondNum = bondNum - len(dummyConnection) - len(
hydrConnection)
newHead = str(newAtomNum).rjust(3) + str(newBondNum).rjust(
3) + fileHead[0][6:]
newLinkerInfoList[fileHeadLineNum] = newHead
newLinkerInfoList[0] = fileName + '\n'
#edit appendix I - MAX NUMBER OF CONTACTS AND ATOMTYPES
contactAppend = []
for i in range(atomNum):
contactAppend.append(
str(contactCount.count(i + 1)) + ' ' + atomTypeList[i])
newLinkerInfoList.append('\n')
newLinkerInfoList.append(
'> <MAX-NUMBER-Of-CONTACTS ATOMTYPES> \n')
newLinkerInfoList = newLinkerInfoList + contactAppend[:
newAtomNum]
newLinkerInfoList.append('\n')
newLinkerInfoList.append('$$$$\n')
#remove dummy atoms
for dummyLine in dummyAtomLineList:
newLinkerInfoList.remove(dummyLine)
#remove dummy bonds
for dummyCon in dummyConnection:
newLinkerInfoList.remove(dummyCon)
#remove hydrogen atoms
for hydrLine in hydrAtomLineList:
newLinkerInfoList.remove(hydrLine)
#remove hydrogen bonds
for hydrCon in hydrConnection:
newLinkerInfoList.remove(hydrCon)
#remove M ISO line
fileMISO = list(
filter(lambda x: 'M ISO' in x, newLinkerInfoList))
if len(fileMISO) > 0:
for ISO in fileMISO:
newLinkerInfoList.remove(ISO)
#remove M CHG line
fileMCHG = list(
filter(lambda x: 'M CHG' in x, newLinkerInfoList))
if len(fileMCHG) > 0:
for CHG in fileMCHG:
newLinkerInfoList.remove(CHG)
#write linker info to file
with open(filePath, 'w') as outf:
outf.writelines(newLinkerInfoList)
with open(outputFolderPath_log + 'Process.log', 'at') as outLog:
outLog.write(time.asctime(time.localtime(time.time())))
outLog.write(' CHOP-MOL ')
outLog.write(inputPath)
outLog.write('\n')
tempCombineList = []
tempCombineList.append(inputPath)
tempCombineList = tempCombineList + fileList
combineLinkers(outputDir, tempCombineList)
