def InteractionFingerprint(ligand, protein, strict=True):
"""Interaction fingerprint accomplished by converting the molecular
interaction of ligand-protein into bit array according to
the residue of choice and the interaction. For every residue
(One row = one residue) there are eight bits which represent
eight type of interactions:
- (Column 0) hydrophobic contacts
- (Column 1) aromatic face to face
- (Column 2) aromatic edge to face
- (Column 3) hydrogen bond (protein as hydrogen bond donor)
- (Column 4) hydrogen bond (protein as hydrogen bond acceptor)
- (Column 5) salt bridges (protein positively charged)
- (Column 6) salt bridges (protein negatively charged)
- (Column 7) salt bridges (ionic bond with metal ion)
Parameters
----------
ligand, protein : oddt.toolkit.Molecule object
Molecules, which are analysed in order to find interactions.
strict : bool (deafult = True)
If False, do not include condition, which informs whether atoms
form 'strict' H-bond (pass all angular cutoffs).
Returns
-------
InteractionFingerprint : numpy array
Vector of calculated IFP (size = no residues * 8 type of interaction)
"""
resids = np.unique(protein.atom_dict['resid'])
IFP = np.zeros((len(resids), 8), dtype=np.uint8)
# hydrophobic contacts (column = 0)
hydrophobic = hydrophobic_contacts(protein, ligand)[0]['resid']
np.add.at(IFP, (np.searchsorted(resids, np.sort(hydrophobic)[::-1]), 0), 1)
# aromatic face to face (Column = 1), aromatic edge to face (Column = 2)
rings, _, strict_parallel, strict_perpendicular = pi_stacking(
protein, ligand)
np.add.at(IFP, (np.searchsorted(
resids, np.sort(rings[strict_parallel]['resid'])[::-1]), 1), 1)
np.add.at(IFP, (np.searchsorted(
resids, np.sort(rings[strict_perpendicular]['resid'])[::-1]), 2), 1)
# h-bonds, protein as a donor (Column = 3)
_, donors, strict0 = hbond_acceptor_donor(ligand, protein)
if strict is False:
strict0 = None
np.add.at(IFP, (np.searchsorted(
resids, np.sort(donors[strict0]['resid'])[::-1]), 3), 1)
# h-bonds, protein as an acceptor (Column = 4)
acceptors, _, strict1 = hbond_acceptor_donor(protein, ligand)
if strict is False:
strict1 = None
np.add.at(IFP, (np.searchsorted(
resids, np.sort(acceptors[strict1]['resid'])[::-1]), 4), 1)
# salt bridges, protein positively charged (Column = 5)
plus, _ = salt_bridge_plus_minus(protein, ligand)
np.add.at(IFP, (np.searchsorted(resids, np.sort(plus['resid'])[::-1]), 5), 1)
# salt bridges, protein negatively charged (Colum = 6)
_, minus = salt_bridge_plus_minus(ligand, protein)
np.add.at(IFP, (np.searchsorted(resids, np.sort(minus['resid'])[::-1]), 6), 1)
# salt bridges, ionic bond with metal ion (Column = 7)
_, metal, strict2 = acceptor_metal(protein, ligand)
if strict is False:
strict2 = None
np.add.at(IFP, (np.searchsorted(
resids, np.sort(metal[strict2]['resid'])[::-1]), 7), 1)
return IFP.flatten()
def SimpleInteractionFingerprint(ligand, protein, strict=True):
"""Based on http://dx.doi.org/10.1016/j.csbj.2014.05.004.
Every IFP consists of 8 bits per amino acid (One row = one amino acid)
and present eight type of interaction:
- (Column 0) hydrophobic contacts
- (Column 1) aromatic face to face
- (Column 2) aromatic edge to face
- (Column 3) hydrogen bond (protein as hydrogen bond donor)
- (Column 4) hydrogen bond (protein as hydrogen bond acceptor)
- (Column 5) salt bridges (protein positively charged)
- (Column 6) salt bridges (protein negatively charged)
- (Column 7) salt bridges (ionic bond with metal ion)
Returns matrix, which is sorted according to this pattern : 'ALA',
'ARG', 'ASN', 'ASP', 'CYS', 'GLN', 'GLU', 'GLY', 'HIS', 'ILE', 'LEU',
'LYS', 'MET', 'PHE', 'PRO', 'SER', 'THR', 'TRP', 'TYR', 'VAL', ''.
The '' means cofactor. Index of amino acid in pattern coresponds
to row in returned matrix.
Parameters
----------
ligand, protein : oddt.toolkit.Molecule object
Molecules, which are analysed in order to find interactions.
strict : bool (deafult = True)
If False, do not include condition, which informs whether atoms
form 'strict' H-bond (pass all angular cutoffs).
Returns
-------
InteractionFingerprint : numpy array
Vector of calculated IFP (size = 168)
"""
amino_acids = np.array(['', 'ALA', 'ARG', 'ASN', 'ASP', 'CYS', 'GLN', 'GLU',
'GLY', 'HIS', 'ILE', 'LEU', 'LYS', 'MET', 'PHE',
'PRO', 'SER', 'THR', 'TRP', 'TYR', 'VAL'],
dtype='<U3')
IFP = np.zeros((len(amino_acids), 8), dtype=np.uint8)
# hydrophobic (Column = 0)
hydrophobic = hydrophobic_contacts(protein, ligand)[0]['resname']
hydrophobic[~np.in1d(hydrophobic, amino_acids)] = ''
np.add.at(IFP, (np.searchsorted(amino_acids,
np.sort(hydrophobic)[::-1]), 0), 1)
# aromatic face to face (Column = 1), aromatic edge to face (Column = 2)
rings, _, strict_parallel, strict_perpendicular = pi_stacking(
protein, ligand)
rings[strict_parallel]['resname'][~np.in1d(
rings[strict_parallel]['resname'], amino_acids)] = ''
np.add.at(IFP, (np.searchsorted(
amino_acids, np.sort(rings[strict_parallel]['resname'])[::-1]), 1), 1)
rings[strict_perpendicular]['resname'][~np.in1d(
rings[strict_perpendicular]['resname'], amino_acids)] = ''
np.add.at(IFP, (np.searchsorted(
amino_acids,
np.sort(rings[strict_perpendicular]['resname'])[::-1]), 2), 1)
# hbonds donated by the protein (Column = 3)
_, donors, strict0 = hbond_acceptor_donor(ligand, protein)
donors['resname'][~np.in1d(donors['resname'], amino_acids)] = ''
if strict is False:
strict0 = None
np.add.at(IFP, (np.searchsorted(
amino_acids, np.sort(donors[strict0]['resname'])[::-1]), 3), 1)
# hbonds donated by the ligand (Column = 4)
acceptors, _, strict1 = hbond_acceptor_donor(protein, ligand)
acceptors['resname'][~np.in1d(acceptors['resname'], amino_acids)] = ''
if strict is False:
strict1 = None
np.add.at(IFP, (np.searchsorted(
amino_acids, np.sort(acceptors[strict1]['resname'])[::-1]), 4), 1)
# ionic bond with protein cation(Column = 5)
plus, _ = salt_bridge_plus_minus(protein, ligand)
plus['resname'][~np.in1d(plus['resname'], amino_acids)] = ''
np.add.at(IFP, (np.searchsorted(amino_acids,
np.sort(plus['resname'])[::-1]), 5), 1)
# ionic bond with protein anion(Column = 6)
_, minus = salt_bridge_plus_minus(ligand, protein)
minus['resname'][~np.in1d(minus['resname'], amino_acids)] = ''
np.add.at(IFP, (np.searchsorted(amino_acids,
np.sort(minus['resname'])[::-1]), 6), 1)
# ionic bond with metal ion (Column = 7)
_, metal, strict2 = acceptor_metal(protein, ligand)
metal['resname'][~np.in1d(metal['resname'], amino_acids)] = ''
if strict is False:
strict2 = None
np.add.at(IFP, (np.searchsorted(
amino_acids, np.sort(metal[strict2]['resname'])[::-1]), 7), 1)
return IFP.flatten()
def SimpleInteractionFingerprint(ligand, protein, strict=True):
"""Based on http://dx.doi.org/10.1016/j.csbj.2014.05.004.
Every IFP consists of 8 bits per amino acid (One row = one amino acid)
and present eight type of interaction:
- (Column 0) hydrophobic contacts
- (Column 1) aromatic face to face
- (Column 2) aromatic edge to face
- (Column 3) hydrogen bond (protein as hydrogen bond donor)
- (Column 4) hydrogen bond (protein as hydrogen bond acceptor)
- (Column 5) salt bridges (protein positively charged)
- (Column 6) salt bridges (protein negatively charged)
- (Column 7) salt bridges (ionic bond with metal ion)
Returns matrix, which is sorted according to this pattern : 'ALA',
'ARG', 'ASN', 'ASP', 'CYS', 'GLN', 'GLU', 'GLY', 'HIS', 'ILE', 'LEU',
'LYS', 'MET', 'PHE', 'PRO', 'SER', 'THR', 'TRP', 'TYR', 'VAL', ''.
The '' means cofactor. Index of amino acid in pattern coresponds
to row in returned matrix.
Parameters
----------
ligand, protein : oddt.toolkit.Molecule object
Molecules, which are analysed in order to find interactions.
strict : bool (deafult = True)
If False, do not include condition, which informs whether atoms
form 'strict' H-bond (pass all angular cutoffs).
Returns
-------
InteractionFingerprint : numpy array
Vector of calculated IFP (size = 168)
"""
amino_acids = np.array(['', 'ALA', 'ARG', 'ASN', 'ASP', 'CYS', 'GLN', 'GLU',
'GLY', 'HIS', 'ILE', 'LEU', 'LYS', 'MET', 'PHE',
'PRO', 'SER', 'THR', 'TRP', 'TYR', 'VAL'],
dtype='<U3')
IFP = np.zeros((len(amino_acids), 8), dtype=np.uint8)
# hydrophobic (Column = 0)
hydrophobic = hydrophobic_contacts(protein, ligand)[0]['resname']
hydrophobic[~np.in1d(hydrophobic, amino_acids)] = ''
np.add.at(IFP, [np.searchsorted(amino_acids,
np.sort(hydrophobic)[::-1]), 0], 1)
# aromatic face to face (Column = 1), aromatic edge to face (Column = 2)
rings, _, strict_parallel, strict_perpendicular = pi_stacking(
protein, ligand)
rings[strict_parallel]['resname'][~np.in1d(
rings[strict_parallel]['resname'], amino_acids)] = ''
np.add.at(IFP, [np.searchsorted(
amino_acids, np.sort(rings[strict_parallel]['resname'])[::-1]), 1], 1)
rings[strict_perpendicular]['resname'][~np.in1d(
rings[strict_perpendicular]['resname'], amino_acids)] = ''
np.add.at(IFP, [np.searchsorted(
amino_acids,
np.sort(rings[strict_perpendicular]['resname'])[::-1]), 2], 1)
# hbonds donated by the protein (Column = 3)
_, donors, strict0 = hbond_acceptor_donor(ligand, protein)
donors['resname'][~np.in1d(donors['resname'], amino_acids)] = ''
if strict is False:
strict0 = None
np.add.at(IFP, [np.searchsorted(
amino_acids, np.sort(donors[strict0]['resname'])[::-1]), 3], 1)
# hbonds donated by the ligand (Column = 4)
acceptors, _, strict1 = hbond_acceptor_donor(protein, ligand)
acceptors['resname'][~np.in1d(acceptors['resname'], amino_acids)] = ''
if strict is False:
strict1 = None
np.add.at(IFP, [np.searchsorted(
amino_acids, np.sort(acceptors[strict1]['resname'])[::-1]), 4], 1)
# ionic bond with protein cation(Column = 5)
plus, _ = salt_bridge_plus_minus(protein, ligand)
plus['resname'][~np.in1d(plus['resname'], amino_acids)] = ''
np.add.at(IFP, [np.searchsorted(amino_acids,
np.sort(plus['resname'])[::-1]), 5], 1)
# ionic bond with protein anion(Column = 6)
_, minus = salt_bridge_plus_minus(ligand, protein)
minus['resname'][~np.in1d(minus['resname'], amino_acids)] = ''
np.add.at(IFP, [np.searchsorted(amino_acids,
np.sort(minus['resname'])[::-1]), 6], 1)
# ionic bond with metal ion (Column = 7)
_, metal, strict2 = acceptor_metal(protein, ligand)
metal['resname'][~np.in1d(metal['resname'], amino_acids)] = ''
if strict is False:
strict2 = None
np.add.at(IFP, [np.searchsorted(
amino_acids, np.sort(metal[strict2]['resname'])[::-1]), 7], 1)
return IFP.flatten()
def InteractionFingerprint(ligand, protein, strict=True):
"""Interaction fingerprint accomplished by converting the molecular
interaction of ligand-protein into bit array according to
the residue of choice and the interaction. For every residue
(One row = one residue) there are eight bits which represent
eight type of interactions:
- (Column 0) hydrophobic contacts
- (Column 1) aromatic face to face
- (Column 2) aromatic edge to face
- (Column 3) hydrogen bond (protein as hydrogen bond donor)
- (Column 4) hydrogen bond (protein as hydrogen bond acceptor)
- (Column 5) salt bridges (protein positively charged)
- (Column 6) salt bridges (protein negatively charged)
- (Column 7) salt bridges (ionic bond with metal ion)
Parameters
----------
ligand, protein : oddt.toolkit.Molecule object
Molecules, which are analysed in order to find interactions.
strict : bool (deafult = True)
If False, do not include condition, which informs whether atoms
form 'strict' H-bond (pass all angular cutoffs).
Returns
-------
InteractionFingerprint : numpy array
Vector of calculated IFP (size = no residues * 8 type of interaction)
"""
resids = np.unique(protein.atom_dict['resid'])
IFP = np.zeros((len(resids), 8), dtype=np.uint8)
# hydrophobic contacts (column = 0)
hydrophobic = hydrophobic_contacts(protein, ligand)[0]['resid']
np.add.at(IFP, [np.searchsorted(resids, np.sort(hydrophobic)[::-1]), 0], 1)
# aromatic face to face (Column = 1), aromatic edge to face (Column = 2)
rings, _, strict_parallel, strict_perpendicular = pi_stacking(
protein, ligand)
np.add.at(IFP, [np.searchsorted(
resids, np.sort(rings[strict_parallel]['resid'])[::-1]), 1], 1)
np.add.at(IFP, [np.searchsorted(
resids, np.sort(rings[strict_perpendicular]['resid'])[::-1]), 2], 1)
# h-bonds, protein as a donor (Column = 3)
_, donors, strict0 = hbond_acceptor_donor(ligand, protein)
if strict is False:
strict0 = None
np.add.at(IFP, [np.searchsorted(
resids, np.sort(donors[strict0]['resid'])[::-1]), 3], 1)
# h-bonds, protein as an acceptor (Column = 4)
acceptors, _, strict1 = hbond_acceptor_donor(protein, ligand)
if strict is False:
strict1 = None
np.add.at(IFP, [np.searchsorted(
resids, np.sort(acceptors[strict1]['resid'])[::-1]), 4], 1)
# salt bridges, protein positively charged (Column = 5)
plus, _ = salt_bridge_plus_minus(protein, ligand)
np.add.at(IFP, [np.searchsorted(resids, np.sort(plus['resid'])[::-1]), 5], 1)
# salt bridges, protein negatively charged (Colum = 6)
_, minus = salt_bridge_plus_minus(ligand, protein)
np.add.at(IFP, [np.searchsorted(resids, np.sort(minus['resid'])[::-1]), 6], 1)
# salt bridges, ionic bond with metal ion (Column = 7)
_, metal, strict2 = acceptor_metal(protein, ligand)
if strict is False:
strict2 = None
np.add.at(IFP, [np.searchsorted(
resids, np.sort(metal[strict2]['resid'])[::-1]), 7], 1)
return IFP.flatten()
def InteractionCheck(ppath, Listoflig, cur_dir):
global proteinpath
proteinpath = ppath
os.chdir(os.path.dirname(proteinpath))
# pname = os.path.basename(proteinpath)
# protein = next(oddt.toolkit.readfile('pdb', proteinpath, removeHs=False, cleanupSubstructures=False, sanitize=False))
try:
protein = next(oddt.toolkit.readfile('pdb', proteinpath, removeHs=False))
protein.protein = True
except Exception as e:
print("Input structure could not be split into protein and ligand. Please check ligand identifier.")
f2 = open(os.path.join(os.path.basename(proteinpath), 'ErrorLog.txt'), 'w')
f2.write(str(e))
f2.close()
for ligand_object in Listoflig:
ligandname = ligand_object.PoseNameExt
ResReport = ligand_object.PoseName + "_ResidueReport.csv"
path = os.path.join(cur_dir, 'Fingerprint', ResReport)
file = open(path, 'w')
file.write("Ligand interactions with protein residues\n")
file.close()
# Read in and define the reference ligand
ligand = next(oddt.toolkit.readfile('pdb', ligandname, removeHs=False))
# Hydrophobic interactions
p_hydroph, l_hydroph = interactions.hydrophobic_contacts(protein, ligand)
InteractionsFile(p_hydroph, l_hydroph, path, 'hydrophobic')
# h bonds
p_hbonds, l_hbonds, strict = interactions.hbonds(protein, ligand)
InteractionsFile(p_hbonds, l_hbonds, path, 'hydrogen bond')
# halogens
p_halogen, l_halogen, strict = interactions.halogenbonds(protein, ligand)
InteractionsFile(p_halogen, l_halogen, path, 'halogen bond')
# pistacking bonds
pi_interactions = interactions.pi_stacking(protein, ligand)
InteractionsFile(pi_interactions[0], pi_interactions[2], path, 'pi stacking')
# salt bridges
p_salt_bridges, l_salt_bridges = interactions.salt_bridges(protein, ligand)
InteractionsFile(p_salt_bridges, l_salt_bridges, path, 'salt bridge')
# pi_cation
p_pi_cation, l_pi_cation, strict = interactions.pi_cation(protein, ligand)
InteractionsFile(p_pi_cation, l_pi_cation, path, 'pi cation')
# acceptor_metal bonds
p_acceptor_metal_a, acceptor_metal_a, strict = interactions.acceptor_metal(protein, ligand)
InteractionsFile(p_acceptor_metal_a, acceptor_metal_a, path, 'acceptor metal')
# pi_metal bonds
p_pi_metal, l_pi_metal, strict = interactions.pi_metal(protein, ligand)
InteractionsFile(p_pi_metal, l_pi_metal, path, 'pi metal')
