def toMolecule(self, formalcharges=False, ids=None):
"""
Return the moleculekit.molecule.Molecule
Parameters
----------
formalcharges: bool
If True,the formal charges are used instead of partial ones
ids: list
The list of conformer ids to store in the moleculekit Molecule object- If None, all are returned
Default: None
Returns
-------
mol: moleculekit.molecule.Molecule
The moleculekit Molecule object
"""
from moleculekit.molecule import Molecule
class NoConformerError(Exception):
pass
if ids is None:
ids = np.arange(self.numFrames)
if self.numFrames == 0:
raise NoConformerError(
"No Conformers are found in the molecule. Generate at least one conformer."
)
elif not isinstance(ids, list) and not isinstance(ids, np.ndarray):
raise ValueError(
'The argument ids should be a list of confomer ids')
mol = Molecule()
mol.empty(self.numAtoms)
mol.record[:] = 'HETATM'
mol.resname[:] = self.ligname[:3]
mol.resid[:] = self._resid
mol.coords = self._coords[:, :, ids]
mol.name[:] = self._name
mol.element[:] = self._element
if formalcharges:
mol.charge[:] = self._formalcharge
else:
mol.charge[:] = self._charge
mol.box = np.zeros((3, self.numFrames), dtype=np.float32)
mol.viewname = self.ligname
mol.bonds = self._bonds
mol.bondtype = self._bondtype
mol.atomtype = self._atomtype
return mol
def setUpClass(self):
from moleculekit.home import home
from moleculekit.molecule import Molecule
from moleculekit.tools.preparation import proteinPrepare
from moleculekit.tools.autosegment import autoSegment
self.testf = os.path.join(home(), 'test-data', 'test-voxeldescriptors')
mol = Molecule(os.path.join(self.testf, '3PTB.pdb'))
mol.filter('protein')
mol = autoSegment(mol, field='both')
mol = proteinPrepare(mol)
mol.bonds = mol._guessBonds()
self.mol = mol
def setUpClass(self):
from moleculekit.home import home
from moleculekit.molecule import Molecule
from moleculekit.tools.preparation import systemPrepare
from moleculekit.tools.autosegment import autoSegment
self.testf = os.path.join(home(), "test-data", "test-voxeldescriptors")
mol = Molecule(os.path.join(self.testf, "3PTB.pdb"))
mol.filter("protein")
mol = autoSegment(mol, field="both")
mol = systemPrepare(mol, pH=7.0)
mol.bonds = mol._guessBonds()
self.mol = mol
def test_metricsphericalcoordinate(self):
from moleculekit.molecule import Molecule
from moleculekit.home import home
from os import path
mol = Molecule(
path.join(home(dataDir='test-projections'), 'trajectory',
'filtered.pdb'))
ref = mol.copy()
mol.read(
path.join(home(dataDir='test-projections'), 'trajectory',
'traj.xtc'))
mol.bonds = mol._guessBonds()
res = MetricSphericalCoordinate(ref, 'resname MOL',
'within 8 of resid 98').project(mol)
_ = MetricSphericalCoordinate(ref, 'resname MOL',
'within 8 of resid 98').getMapping(mol)
ref_array = np.load(
path.join(home(dataDir='test-projections'),
'metricsphericalcoordinate', 'res.npy'))
assert np.allclose(res, ref_array, rtol=0, atol=1e-04)
def pdb2psf_CA(pdb_name_in, psf_name_out, bonds=True, angles=True):
mol = Molecule(pdb_name_in)
mol.filter("name CA")
n = mol.numAtoms
atom_types = []
for i in range(n):
atom_types.append(CA_MAP[(mol.resname[i], mol.name[i])])
if bonds:
bonds = np.concatenate(
(
np.arange(n - 1).reshape([n - 1, 1]),
(np.arange(1, n).reshape([n - 1, 1])),
),
axis=1,
)
else:
bonds = np.empty([0, 2], dtype=np.int32)
if angles:
angles = np.concatenate(
(
np.arange(n - 2).reshape([n - 2, 1]),
(np.arange(1, n - 1).reshape([n - 2, 1])),
(np.arange(2, n).reshape([n - 2, 1])),
),
axis=1,
)
else:
angles = np.empty([0, 3], dtype=np.int32)
mol.atomtype = np.array(atom_types)
mol.bonds = bonds
mol.angles = angles
mol.write(psf_name_out)
def build(mol,
ff=None,
topo=None,
param=None,
prefix='structure',
outdir='./build',
caps=None,
ionize=True,
saltconc=0,
saltanion=None,
saltcation=None,
disulfide=None,
teleap=None,
teleapimports=None,
execute=True,
atomtypes=None,
offlibraries=None,
gbsa=False,
igb=2):
""" Builds a system for AMBER
Uses tleap to build a system for AMBER. Additionally it allows the user to ionize and add disulfide bridges.
Parameters
----------
mol : :class:`Molecule <moleculekit.molecule.Molecule>` object
The Molecule object containing the system
ff : list of str
A list of leaprc forcefield files.
Use :func:`amber.listFiles <htmd.builder.amber.listFiles>` to get a list of available forcefield files.
Default: :func:`amber.defaultFf <htmd.builder.amber.defaultFf>`
topo : list of str
A list of topology `prepi/prep/in` files.
Use :func:`amber.listFiles <htmd.builder.amber.listFiles>` to get a list of available topology files.
Default: :func:`amber.defaultTopo <htmd.builder.amber.defaultTopo>`
param : list of str
A list of parameter `frcmod` files.
Use :func:`amber.listFiles <htmd.builder.amber.listFiles>` to get a list of available parameter files.
Default: :func:`amber.defaultParam <htmd.builder.amber.defaultParam>`
prefix : str
The prefix for the generated pdb and psf files
outdir : str
The path to the output directory
Default: './build'
caps : dict
A dictionary with keys segids and values lists of strings describing the caps for a particular protein segment.
e.g. caps['P'] = ['ACE', 'NME'] or caps['P'] = ['none', 'none']. Default: will apply ACE and NME caps to every
protein segment.
ionize : bool
Enable or disable ionization
saltconc : float
Salt concentration to add to the system after neutralization.
saltanion : {'Cl-'}
The anion type. Please use only AMBER ion atom names.
saltcation : {'Na+', 'K+', 'Cs+'}
The cation type. Please use only AMBER ion atom names.
disulfide : list of pairs of atomselection strings
If None it will guess disulfide bonds. Otherwise provide a list pairs of atomselection strings for each pair of
residues forming the disulfide bridge.
teleap : str
Path to teLeap executable used to build the system for AMBER
teleapimports : list
A list of paths to pass to teLeap '-I' flag, i.e. directories to be searched
Default: determined from :func:`amber.defaultAmberHome <htmd.builder.amber.defaultAmberHome>` and
:func:`amber.htmdAmberHome <htmd.builder.amber.htmdAmberHome>`
execute : bool
Disable building. Will only write out the input script needed by tleap. Does not include ionization.
atomtypes : list of triplets
Custom atom types defined by the user as ('type', 'element', 'hybrid') triplets
e.g. (('C1', 'C', 'sp2'), ('CI', 'C', 'sp3')). Check `addAtomTypes` in AmberTools docs.
offlibraries : str or list
A path or a list of paths to OFF library files. Check `loadOFF` in AmberTools docs.
gbsa : bool
Modify radii for GBSA implicit water model
igb : int
GB model. Select: 1 for mbondi, 2 and 5 for mbondi2, 7 for bondi and 8 for mbondi3.
Check section 4. The Generalized Born/Surface Area Model of the AMBER manual.
Returns
-------
molbuilt : :class:`Molecule <moleculekit.molecule.Molecule>` object
The built system in a Molecule object
Example
-------
>>> from htmd.ui import * # doctest: +SKIP
>>> mol = Molecule("3PTB")
>>> molbuilt = amber.build(mol, outdir='/tmp/build') # doctest: +SKIP
>>> # More complex example
>>> disu = [['segid P and resid 157', 'segid P and resid 13'], ['segid K and resid 1', 'segid K and resid 25']]
>>> molbuilt = amber.build(mol, outdir='/tmp/build', saltconc=0.15, disulfide=disu) # doctest: +SKIP
"""
# Remove pdb protein bonds as they can be regenerated by tleap. Keep non-protein bonds i.e. for ligands
mol = mol.copy()
_removeProteinBonds(mol)
if teleap is None:
teleap = _findTeLeap()
else:
if shutil.which(teleap) is None:
raise NameError(
'Could not find executable: `{}` in the PATH. Cannot build for AMBER.'
.format(teleap))
if not os.path.isdir(outdir):
os.makedirs(outdir)
_cleanOutDir(outdir)
if ff is None:
ff = defaultFf()
if topo is None:
topo = defaultTopo()
if param is None:
param = defaultParam()
if caps is None:
caps = _defaultProteinCaps(mol)
_missingSegID(mol)
_checkMixedSegment(mol)
mol = _charmmLipid2Amber(mol)
_applyProteinCaps(mol, caps)
f = open(os.path.join(outdir, 'tleap.in'), 'w')
f.write('# tleap file generated by amber.build\n')
# Printing out the forcefields
for i, force in enumerate(ensurelist(ff)):
if not os.path.isfile(force):
force = _locateFile(force, 'ff', teleap)
if force is None:
continue
newname = 'ff{}_{}'.format(i, os.path.basename(force))
shutil.copy(force, os.path.join(outdir, newname))
f.write('source {}\n'.format(newname))
f.write('\n')
if gbsa:
gbmodels = {
1: 'mbondi',
2: 'mbondi2',
5: 'mbondi2',
7: 'bondi',
8: 'mbondi3'
}
f.write('set default PBradii {}\n\n'.format(gbmodels[igb]))
# Adding custom atom types
if atomtypes is not None:
atomtypes = ensurelist(tocheck=atomtypes[0], tomod=atomtypes)
f.write('addAtomTypes {\n')
for at in atomtypes:
if len(at) != 3:
raise RuntimeError(
'Atom type definitions have to be triplets. Check the AMBER documentation.'
)
f.write(' {{ "{}" "{}" "{}" }}\n'.format(at[0], at[1], at[2]))
f.write('}\n\n')
# Loading OFF libraries
if offlibraries is not None:
offlibraries = ensurelist(offlibraries)
for i, off in enumerate(offlibraries):
if not os.path.isfile(off):
raise RuntimeError(
'Could not find off-library in location {}'.format(off))
newname = 'offlib{}_{}'.format(i, os.path.basename(off))
shutil.copy(off, os.path.join(outdir, newname))
f.write('loadoff {}\n'.format(newname))
# Loading frcmod parameters
f.write('# Loading parameter files\n')
for i, p in enumerate(param):
if not os.path.isfile(p):
p = _locateFile(p, 'param', teleap)
if p is None:
continue
newname = 'param{}_{}'.format(i, os.path.basename(p))
shutil.copy(p, os.path.join(outdir, newname))
f.write('loadamberparams {}\n'.format(newname))
f.write('\n')
# Loading prepi topologies
f.write('# Loading prepi topologies\n')
for i, t in enumerate(topo):
if not os.path.isfile(t):
t = _locateFile(t, 'topo', teleap)
if t is None:
continue
newname = 'topo{}_{}'.format(i, os.path.basename(t))
shutil.copy(t, os.path.join(outdir, newname))
f.write('loadamberprep {}\n'.format(newname))
f.write('\n')
f.write('# Loading the system\n')
f.write('mol = loadpdb input.pdb\n\n')
if np.sum(mol.atomtype != '') != 0:
logger.debug('Writing mol2 files for input to tleap.')
segs = np.unique(mol.segid[mol.atomtype != ''])
combstr = 'mol = combine {mol'
for s in segs:
name = 'segment{}'.format(s)
mol2name = os.path.join(outdir, '{}.mol2'.format(name))
mol.write(mol2name, (mol.atomtype != '') & (mol.segid == s))
if not os.path.isfile(mol2name):
raise NameError(
'Could not write a mol2 file out of the given Molecule.')
f.write('# Loading the rest of the system\n')
f.write('{} = loadmol2 {}.mol2\n\n'.format(name, name))
combstr += ' {}'.format(name)
combstr += '}\n\n'
f.write(combstr)
# Write patches for disulfide bonds (only after ionizing)
if not ionize:
# TODO: Remove this once we deprecate the class
from htmd.builder.builder import DisulfideBridge
from moleculekit.molecule import UniqueResidueID
if disulfide is not None and len(disulfide) != 0 and isinstance(
disulfide[0], DisulfideBridge):
newdisu = []
for d in disulfide:
r1 = UniqueResidueID.fromMolecule(
mol, 'resid {} and segname {}'.format(d.resid1, d.segid1))
r2 = UniqueResidueID.fromMolecule(
mol, 'resid {} and segname {}'.format(d.resid2, d.segid2))
newdisu.append([r1, r2])
disulfide = newdisu
# TODO: Remove up to here ----------------------
if disulfide is not None and len(disulfide) != 0 and isinstance(
disulfide[0][0], str):
disulfide = convertDisulfide(mol, disulfide)
if disulfide is None:
logger.info('Detecting disulfide bonds.')
disulfide = detectDisulfideBonds(mol)
# Fix structure to match the disulfide patching
if len(disulfide) != 0:
torem = np.zeros(mol.numAtoms, dtype=bool)
f.write('# Adding disulfide bonds\n')
for d in disulfide:
# Rename the residues to CYX if there is a disulfide bond
atoms1 = d[0].selectAtoms(mol, indexes=False)
atoms2 = d[1].selectAtoms(mol, indexes=False)
mol.resname[atoms1] = 'CYX'
mol.resname[atoms2] = 'CYX'
# Remove (eventual) HG hydrogens on these CYS (from proteinPrepare)
torem |= (atoms1 & (mol.name == 'HG')) | (atoms2 &
(mol.name == 'HG'))
# Convert to stupid amber residue numbering
uqseqid = sequenceID(
(mol.resid, mol.insertion, mol.segid)) + mol.resid[0]
uqres1 = int(np.unique(uqseqid[atoms1]))
uqres2 = int(np.unique(uqseqid[atoms2]))
f.write('bond mol.{}.SG mol.{}.SG\n'.format(uqres1, uqres2))
f.write('\n')
mol.remove(torem, _logger=False)
# Calculate the bounding box and store it in the CRD file
f.write('setBox mol "vdw"\n\n')
f.write('# Writing out the results\n')
f.write('saveamberparm mol ' + prefix + '.prmtop ' + prefix + '.crd\n')
f.write('quit')
f.close()
# Printing and loading the PDB file. AMBER can work with a single PDB file if the segments are separate by TER
logger.debug('Writing PDB file for input to tleap.')
pdbname = os.path.join(outdir, 'input.pdb')
# mol2 files have atomtype, here we only write parts not coming from mol2
# We need to write the input.pdb at the end since we modify the resname for disulfide bridges in mol
mol.write(pdbname, mol.atomtype == '')
if not os.path.isfile(pdbname):
raise NameError(
'Could not write a PDB file out of the given Molecule.')
molbuilt = None
if execute:
if not teleapimports:
teleapimports = []
# Source default Amber (i.e. the same paths tleap imports)
amberhome = defaultAmberHome(teleap=teleap)
teleapimports += [
os.path.join(amberhome, s)
for s in _defaultAmberSearchPaths.values()
]
if len(teleapimports) == 0:
raise RuntimeWarning(
'No default Amber force-field found. Check teLeap location: {}'
.format(teleap))
# Source HTMD Amber paths that contain ffs
htmdamberdir = htmdAmberHome()
teleapimports += [
os.path.join(htmdamberdir, os.path.dirname(f)) for f in ff
if os.path.isfile(os.path.join(htmdamberdir, f))
]
if len(teleapimports) == 0:
raise RuntimeError(
'No default Amber force-field imports found. Check '
'`htmd.builder.amber.defaultAmberHome()` and `htmd.builder.amber.htmdAmberHome()`'
)
# Set import flags for teLeap
teleapimportflags = []
for p in teleapimports:
teleapimportflags.append('-I')
teleapimportflags.append(str(p))
logpath = os.path.abspath(os.path.join(outdir, 'log.txt'))
logger.info('Starting the build.')
currdir = os.getcwd()
os.chdir(outdir)
f = open(logpath, 'w')
try:
cmd = [teleap, '-f', './tleap.in']
cmd[1:1] = teleapimportflags
logger.debug(cmd)
call(cmd, stdout=f)
except:
raise NameError('teLeap failed at execution')
f.close()
errors = _logParser(logpath)
os.chdir(currdir)
if errors:
raise BuildError(errors + [
'Check {} for further information on errors in building.'.
format(logpath)
])
logger.info('Finished building.')
if os.path.exists(os.path.join(outdir, 'structure.crd')) and \
os.path.getsize(os.path.join(outdir, 'structure.crd')) != 0 and \
os.path.getsize(os.path.join(outdir, 'structure.prmtop')) != 0:
try:
molbuilt = Molecule(os.path.join(outdir, 'structure.prmtop'))
molbuilt.read(os.path.join(outdir, 'structure.crd'))
except Exception as e:
raise RuntimeError(
'Failed at reading structure.prmtop/structure.crd due to error: {}'
.format(e))
else:
raise BuildError(
'No structure pdb/prmtop file was generated. Check {} for errors in building.'
.format(logpath))
if ionize:
shutil.move(os.path.join(outdir, 'structure.crd'),
os.path.join(outdir, 'structure.noions.crd'))
shutil.move(os.path.join(outdir, 'structure.prmtop'),
os.path.join(outdir, 'structure.noions.prmtop'))
totalcharge = np.sum(molbuilt.charge)
nwater = np.sum(molbuilt.atomselect('water and noh'))
anion, cation, anionatom, cationatom, nanion, ncation = ionizef(
totalcharge,
nwater,
saltconc=saltconc,
anion=saltanion,
cation=saltcation)
newmol = ionizePlace(mol, anion, cation, anionatom, cationatom,
nanion, ncation)
# Redo the whole build but now with ions included
return build(newmol,
ff=ff,
topo=topo,
param=param,
prefix=prefix,
outdir=outdir,
caps={},
ionize=False,
execute=execute,
saltconc=saltconc,
disulfide=disulfide,
teleap=teleap,
atomtypes=atomtypes,
offlibraries=offlibraries)
tmpbonds = molbuilt.bonds
molbuilt.bonds = [] # Removing the bonds to speed up writing
molbuilt.write(os.path.join(outdir, 'structure.pdb'))
molbuilt.bonds = tmpbonds # Restoring the bonds
detectCisPeptideBonds(molbuilt) # Warn in case of cis bonds
return molbuilt
def pdb2psf_CACB(pdb_name_in, psf_name_out, bonds=True, angles=True):
mol = Molecule(pdb_name_in)
mol.filter("name CA CB")
n = mol.numAtoms
atom_types = []
for i in range(n):
atom_types.append(CACB_MAP[(mol.resname[i], mol.name[i])])
CA_idx = []
CB_idx = []
for i, name in enumerate(mol.name):
if name[:2] == "CA":
CA_idx.append(i)
else:
CB_idx.append(i)
if bonds:
CA_bonds = np.concatenate(
(
np.array(CA_idx)[:-1].reshape([len(CA_idx) - 1, 1]),
np.array(CA_idx)[1:].reshape([len(CA_idx) - 1, 1]),
),
axis=1,
)
CB_bonds = np.concatenate(
(
np.array(CB_idx).reshape(len(CB_idx), 1) - 1,
np.array(CB_idx).reshape(len(CB_idx), 1),
),
axis=1,
)
bonds = np.concatenate((CA_bonds, CB_bonds))
else:
bonds = np.empty([0, 2], dtype=np.int32)
if angles:
CA_angles = np.concatenate(
(
np.array(CA_idx)[:-2].reshape([len(CA_idx) - 2, 1]),
np.array(CA_idx)[1:-1].reshape([len(CA_idx) - 2, 1]),
np.array(CA_idx)[2:].reshape([len(CA_idx) - 2, 1]),
),
axis=1,
)
CB_angles = []
cbn = 0
for i in CA_idx:
if mol.resname[i] != "GLY":
if i != 0:
CB_angles.append(
[CA_idx[CA_idx.index(i) - 1], i, CB_idx[cbn]])
if i != CA_idx[-1]:
CB_angles.append(
[CA_idx[CA_idx.index(i) + 1], i, CB_idx[cbn]])
cbn += 1
CB_angles = np.array(CB_angles)
angles = np.concatenate((CA_angles, CB_angles))
else:
angles = np.empty([0, 3], dtype=np.int32)
mol.atomtype = np.array(atom_types)
mol.bonds = bonds
mol.angles = angles
mol.write(psf_name_out)