def testUpdateDataFile(self):
"""Test case - update data file
"""
self.lfh.write("\nStarting %s %s\n" % (self.__class__.__name__,
sys._getframe().f_code.co_name))
try:
# Create a initial data file --
#
myDataList=[]
curContainer=DataContainer("myblock")
aCat=DataCategory("pdbx_seqtool_mapping_ref")
aCat.appendAttribute("ordinal")
aCat.appendAttribute("entity_id")
aCat.appendAttribute("auth_mon_id")
aCat.appendAttribute("auth_mon_num")
aCat.appendAttribute("pdb_chain_id")
aCat.appendAttribute("ref_mon_id")
aCat.appendAttribute("ref_mon_num")
aCat.append([9,2,3,4,5,6,7])
aCat.append([10,2,3,4,5,6,7])
aCat.append([11,2,3,4,5,6,7])
aCat.append([12,2,3,4,5,6,7])
#self.lfh.write("Assigned data category state-----------------\n")
#aCat.dumpIt(fh=self.lfh)
tmpf = tempname(suffix='.cif')
curContainer.append(aCat)
myDataList.append(curContainer)
ofh = open(tmpf, "w")
pdbxW=PdbxWriter(ofh)
pdbxW.write(myDataList)
ofh.close()
#
#
# Read and update the data -
#
myDataList=[]
ifh = open(tmpf, "r")
pRd=PdbxReader(ifh)
pRd.read(myDataList)
ifh.close()
#
myBlock=myDataList[0]
myBlock.printIt()
myCat=myBlock.getObj('pdbx_seqtool_mapping_ref')
myCat.printIt()
for iRow in range(0,myCat.getRowCount()):
myCat.setValue('some value', 'ref_mon_id',iRow)
myCat.setValue(100, 'ref_mon_num',iRow)
ofh = open(tempname(suffix='.cif'), "w")
pdbxW=PdbxWriter(ofh)
pdbxW.write(myDataList)
ofh.close()
#
except:
traceback.print_exc(file=self.lfh)
self.fail()
def MDTRAJwrite(mol, filename):
try:
import mdtraj as md
from htmd.util import tempname
ext = os.path.splitext(filename)[1][1:]
if ext == 'gz':
pieces = filename.split('.')
ext = '{}.{}'.format(pieces[-2], pieces[-1])
if ext in _MDTRAJ_TOPOLOGY_SAVERS:
tmppdb = tempname(suffix='.pdb')
mol.write(tmppdb)
traj = md.load(tmppdb)
os.remove(tmppdb)
elif ext in _MDTRAJ_TRAJECTORY_SAVERS:
tmppdb = tempname(suffix='.pdb')
tmpxtc = tempname(suffix='.xtc')
mol.write(tmppdb)
mol.write(tmpxtc)
traj = md.load(tmpxtc, top=tmppdb)
os.remove(tmppdb)
os.remove(tmpxtc)
else:
raise ValueError('Unknown file type for file {}'.format(filename))
# traj.xyz = np.swapaxes(np.swapaxes(self.coords, 1, 2), 0, 1) / 10
# traj.time = self.time
# traj.unitcell_lengths = self.box.T / 10
traj.save(filename)
except Exception as e:
raise ValueError('MDtraj reader failed for file {} with error "{}"'.format(filename, e))
def MDTRAJwrite(mol, filename):
try:
import mdtraj as md
from htmd.util import tempname
ext = os.path.splitext(filename)[1][1:]
if ext == 'gz':
pieces = filename.split('.')
ext = '{}.{}'.format(pieces[-2], pieces[-1])
if ext in _MDTRAJ_TOPOLOGY_SAVERS:
tmppdb = tempname(suffix='.pdb')
mol.write(tmppdb)
traj = md.load(tmppdb)
os.remove(tmppdb)
elif ext in _MDTRAJ_TRAJECTORY_SAVERS:
tmppdb = tempname(suffix='.pdb')
tmpxtc = tempname(suffix='.xtc')
mol.write(tmppdb)
mol.write(tmpxtc)
traj = md.load(tmpxtc, top=tmppdb)
os.remove(tmppdb)
os.remove(tmpxtc)
else:
raise ValueError('Unknown file type for file {}'.format(filename))
# traj.xyz = np.swapaxes(np.swapaxes(self.coords, 1, 2), 0, 1) / 10
# traj.time = self.time
# traj.unitcell_lengths = self.box.T / 10
traj.save(filename)
except Exception as e:
raise ValueError(
'MDtraj reader failed for file {} with error "{}"'.format(
filename, e))
def testUpdateDataFile(self):
"""Test case - write data file
"""
self.lfh.write("\nStarting %s %s\n" % (self.__class__.__name__,
sys._getframe().f_code.co_name))
try:
# Create a initial data file --
#
tmpf = tempname(suffix='.cif')
myDataList=[]
ofh = open(tmpf, "w")
curContainer=DataContainer("myblock")
aCat=DataCategory("pdbx_seqtool_mapping_ref")
aCat.appendAttribute("ordinal")
aCat.appendAttribute("entity_id")
aCat.appendAttribute("auth_mon_id")
aCat.appendAttribute("auth_mon_num")
aCat.appendAttribute("pdb_chain_id")
aCat.appendAttribute("ref_mon_id")
aCat.appendAttribute("ref_mon_num")
aCat.append((1,2,3,4,5,6,7))
aCat.append((1,2,3,4,5,6,7))
aCat.append((1,2,3,4,5,6,7))
aCat.append((1,2,3,4,5,6,7))
curContainer.append(aCat)
myDataList.append(curContainer)
pdbxW=PdbxWriter(ofh)
pdbxW.write(myDataList)
ofh.close()
#
# Read and update the data -
#
myDataList=[]
ifh = open(tmpf, "r")
pRd=PdbxReader(ifh)
pRd.read(myDataList)
ifh.close()
#
myBlock=myDataList[0]
myBlock.printIt()
myCat=myBlock.getObj('pdbx_seqtool_mapping_ref')
myCat.printIt()
for iRow in range(0,myCat.getRowCount()):
myCat.setValue('some value', 'ref_mon_id',iRow)
myCat.setValue(100, 'ref_mon_num',iRow)
ofh = open(tempname(suffix='.cif'), "w")
pdbxW=PdbxWriter(ofh)
pdbxW.write(myDataList)
ofh.close()
except:
traceback.print_exc(file=sys.stderr)
self.fail()
def test_customDisulfideBonds(self):
from htmd.molecule.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir='test-proteinprepare')
pdb = '1GZM'
inFile = os.path.join(preparedInputDir, pdb,
"{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter('protein') # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ['top/top_all36_prot.rtf', 'top/top_water_ions.rtf']
params = ['par/par_all36_prot_mod.prm', 'par/par_water_ions.prm']
disu = [['segid 1 and resid 110', 'segid 1 and resid 187'],
['segid 0 and resid 110', 'segid 0 and resid 187']]
tmpdir = tempname()
_ = build(smol,
topo=topos,
param=params,
outdir=tmpdir,
disulfide=disu)
compareDir = home(dataDir=os.path.join('test-charmm-build', pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
# TODO: Remove this after deprecation
from htmd.builder.builder import DisulfideBridge
np.random.seed(1) # Needed for ions
disu = [
DisulfideBridge('1', 110, '1', 187),
DisulfideBridge('0', 110, '0', 187)
]
tmpdir = tempname()
_ = build(smol,
topo=topos,
param=params,
outdir=tmpdir,
disulfide=disu)
compareDir = home(dataDir=os.path.join('test-charmm-build', pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
def test_h2o2_dihed_fix(self):
refDir = os.path.join(self.dataDir, 'h2o2_dihed_fix')
self._test(
refDir, tempname(),
'parameterize -m input.mol2 -f GAFF2 --no-min --no-esp --no-geomopt'
)
def test_equilibration(self):
from htmd.home import home
import filecmp
from htmd.util import tempname
from glob import glob
from htmd.molecule.molecule import Molecule
tmpdir = tempname()
pdbid = '3PTB'
builddir = home(dataDir=os.path.join('test-acemd', pdbid, 'build'))
equil = Acemd('equilibration')
mol = Molecule(os.path.join(builddir, 'structure.pdb'))
celldim = mol.coords.max(axis=0) - mol.coords.min(axis=0)
equil.celldimension = ' '.join(['{:3.1f}'.format(val) for val in celldim.squeeze()])
equil.run = '2000'
equil.trajectoryfreq = 200
equil.temperature = 300
equil.write(builddir, tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join('test-acemd', pdbid, 'equil'))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
match, mismatch, error = filecmp.cmpfiles(refdir, tmpdir, files, shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError('Different results produced by Acemd equilibration for '
'test {} between {} and {} in files {}.'.format(pdbid, refdir, tmpdir, mismatch))
def test_acemd3(self):
from htmd.util import tempname
from htmd.home import home
from glob import glob
import os
pd = Production()
pd.runtime = 4
pd.timeunits = "ns"
pd.temperature = 300
pd.useconstraints = True
pd.constraints = {
"protein and name CA": 1,
"protein and noh and not name CA": 0.1,
}
pd.fb_reference = "protein and name CA"
pd.fb_selection = "resname MOL and noh"
pd.fb_box = [-21, 21, -19, 19, 29, 30]
pd.fb_k = 5
tmpdir = tempname()
pd.write(
home(dataDir=os.path.join("test-protocols", "equilibration",
"acemd3", "protLig", "postrun")),
tmpdir,
)
# Compare with reference
refdir = home(dataDir=os.path.join("test-protocols", "production",
"acemd3", "protLig", "prerun"))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, "*"))]
self._compareResultFolders(refdir, tmpdir, "protLig")
def test_build(self):
from htmd.molecule.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir='test-proteinprepare')
pdbids = ['3PTB', '1A25', '1GZM', '1U5U']
for pdb in pdbids:
with self.subTest(pdb=pdb):
print('Building {}'.format(pdb))
inFile = os.path.join(preparedInputDir, pdb,
"{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter(
'protein') # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ['top/top_all36_prot.rtf', 'top/top_water_ions.rtf']
params = [
'par/par_all36_prot_mod.prm', 'par/par_water_ions.prm'
]
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir)
compareDir = home(
dataDir=os.path.join('test-charmm-build', pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
def test_acemd3(self):
from htmd.util import tempname
from htmd.home import home
from glob import glob
import os
eq = Equilibration(_version=3)
eq.runtime = 4
eq.timeunits = 'ns'
eq.temperature = 300
# keep protein on the same place during all equilibration
from htmd.units import convert
eq.constraintsteps = convert(eq.timeunits,
'timesteps',
eq.runtime,
timestep=eq.acemd.timestep)
eq.constraints = {
'protein and name CA': 1,
'protein and noh and not name CA': 0.1
}
eq.fb_reference = 'protein and name CA'
eq.fb_selection = 'resname MOL and noh'
eq.fb_box = [-21, 21, -19, 19, 29, 30]
eq.fb_k = 5
tmpdir = tempname()
eq.write(
home(dataDir=os.path.join('test-protocols', 'build', 'protLig')),
tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join('test-protocols', 'equilibration',
'acemd3', 'protLig', 'prerun'))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
self._compareResultFolders(refdir, tmpdir, 'protLig')
def setUp(self):
if os.environ.get('TRAVIS_OS_NAME') == 'osx':
self.skipTest('Mac does not work!')
self.maxDiff = None
self.dataDir = home(dataDir='test-param')
self.testDir = os.environ.get('TESTDIR', tempname())
def drawIsoSurface(values3d, resolution=1., plot_center=None, viewer=None):
from htmd.vmdviewer import getCurrentViewer
from htmd.molecule.util import writeVoxels
# plot_center should be - molecule.get_center() + 12
if len(values3d.shape) != 3:
raise ValueError("Your provided a box of {} dimensions."
"\nThis only works with dimension of 3".format(
len(values3d.shape)))
from htmd.util import tempname
if viewer is None:
viewer = getCurrentViewer()
mincoor = np.zeros(3, dtype=np.float64)
maxcoor = np.array(values3d.shape, dtype=np.float64)
rescoor = np.array([resolution] * 3)
# Adjust the plotting center
if plot_center is None:
plot_center = maxcoor / 2.
else:
plot_center = np.array(plot_center)
mincoor -= (plot_center + 0.5
) # TODO: Fix so it will work in case resolution != 1.
maxcoor -= (plot_center + 0.5)
outf = tempname(suffix='.cube')
writeVoxels(values3d, outf, mincoor, maxcoor, rescoor)
viewer.send(
'mol new {} type cube first 0 last -1 step 1 waitfor 1 volsets {{0 }}'.
format(outf))
viewer.send('mol modstyle 0 top Isosurface 0.75 0 2 0 1 1')
def test_customDisulfideBonds(self):
from moleculekit.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir="test-proteinprepare")
pdb = "1GZM"
inFile = os.path.join(preparedInputDir, pdb,
"{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter("protein") # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ["top/top_all36_prot.rtf", "top/top_water_ions.rtf"]
params = ["par/par_all36_prot.prm", "par/par_water_ions.prm"]
disu = [
["segid 1 and resid 110", "segid 1 and resid 187"],
["segid 0 and resid 110", "segid 0 and resid 187"],
]
tmpdir = tempname()
_ = build(smol,
topo=topos,
param=params,
outdir=tmpdir,
disulfide=disu)
compareDir = home(dataDir=os.path.join("test-charmm-build", pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
def test_build(self):
from moleculekit.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir="test-proteinprepare")
pdbids = ["3PTB", "1A25", "1GZM", "1U5U"]
for pdb in pdbids:
with self.subTest(pdb=pdb):
print("Building {}".format(pdb))
inFile = os.path.join(preparedInputDir, pdb,
"{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter(
"protein") # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ["top/top_all36_prot.rtf", "top/top_water_ions.rtf"]
params = ["par/par_all36_prot.prm", "par/par_water_ions.prm"]
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir)
compareDir = home(
dataDir=os.path.join("test-charmm-build", pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
def drawIsoSurface(values3d, resolution=1., plot_center=None, viewer=None):
from htmd.vmdviewer import getCurrentViewer
from htmd.molecule.util import writeVoxels
# plot_center should be - molecule.get_center() + 12
if len(values3d.shape) != 3:
raise ValueError("Your provided a box of {} dimensions."
"\nThis only works with dimension of 3".format(len(values3d.shape)))
from htmd.util import tempname
if viewer is None:
viewer = getCurrentViewer()
mincoor = np.zeros(3, dtype=np.float64)
maxcoor = np.array(values3d.shape, dtype=np.float64)
rescoor = np.array([resolution] * 3)
# Adjust the plotting center
if plot_center is None:
plot_center = maxcoor / 2.
else:
plot_center = np.array(plot_center)
mincoor -= (plot_center + 0.5) # TODO: Fix so it will work in case resolution != 1.
maxcoor -= (plot_center + 0.5)
outf = tempname(suffix='.cube')
writeVoxels(values3d, outf, mincoor, maxcoor, rescoor)
viewer.send('mol new {} type cube first 0 last -1 step 1 waitfor 1 volsets {{0 }}'.format(outf))
viewer.send('mol modstyle 0 top Isosurface 0.75 0 2 0 1 1')
def test_run_water(self):
from htmd.util import tempname
from htmd.home import home
from glob import glob
from subprocess import check_output
import subprocess
import shutil
import os
acemd3exe = shutil.which('acemd3', mode=os.X_OK)
if not acemd3exe:
raise NameError('Could not find acemd3, or no execute permissions are given')
for system in ['amber-equil-completed', 'charmm-equil-completed']:
pd = Production(_version=3)
pd.runtime = 5
pd.timeunits = 'steps'
pd.temperature = 300
pd.constraints = {}
# Set these down for tiny box size of water
pd.acemd.cutoff = 3
pd.acemd.switchdistance = 2
######
tmpdir = tempname()
pd.write(home(dataDir=os.path.join('test-acemd', 'tiny-water', system)), tmpdir)
try:
res = check_output(['acemd3', '--platform', 'CPU', os.getenv('ACE3ARG')], cwd=tmpdir)
except subprocess.CalledProcessError as exc:
assert False, f'Failed to run due to error: {exc}\n\n ---> Error log:\n\n{exc.output.decode("ascii")}'
res = res.decode('utf-8').strip()
print(res)
assert res.endswith('Completed simulation!'), 'Failed at system ' + system
def test_benzamidine_cgenff(self):
refDir = os.path.join(self.dataDir, 'benzamidine_cgenff')
self._test(
refDir, tempname(),
'parameterize -m input.mol2 --charge 1 --forcefield CGENFF --no-min --no-esp --no-torsions'
)
def test_disulfideWithInsertion(self):
from htmd.molecule.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir='test-proteinprepare')
pdb = '3PTB'
print('Building {}'.format(pdb))
inFile = os.path.join(preparedInputDir, pdb,
"{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter('protein') # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ['top/top_all36_prot.rtf', 'top/top_water_ions.rtf']
params = ['par/par_all36_prot_mod.prm', 'par/par_water_ions.prm']
smol.insertion[
smol.resid ==
42] = 'A' # Adding an insertion to test that disulfide bonds with insertions work
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir)
compareDir = home(
dataDir=os.path.join('test-charmm-build', '3PTB_insertion'))
assertSameAsReferenceDir(compareDir, tmpdir)
def test_build(self):
from htmd.molecule.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir='test-proteinprepare')
pdbids = ['3PTB', '1A25', '1GZM', '1U5U']
for pdb in pdbids:
with self.subTest(pdb=pdb):
print('Building {}'.format(pdb))
inFile = os.path.join(preparedInputDir, pdb, "{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter('protein') # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ['top/top_all36_prot.rtf', 'top/top_water_ions.rtf']
params = ['par/par_all36_prot_mod.prm', 'par/par_water_ions.prm']
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir)
compareDir = home(dataDir=os.path.join('test-charmm-build', pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
def test_h2o2_outdir(self):
refDir = os.path.join(self.dataDir, 'h2o2_outdir')
self._test(
refDir, tempname(),
'parameterize -m input.mol2 -f GAFF2 --no-min --no-esp --no-torsions -o dir'
)
def test_disulfideWithInsertion(self):
from htmd.molecule.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir='test-proteinprepare')
pdb = '3PTB'
print('Building {}'.format(pdb))
inFile = os.path.join(preparedInputDir, pdb, "{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter('protein') # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ['top/top_all36_prot.rtf', 'top/top_water_ions.rtf']
params = ['par/par_all36_prot_mod.prm', 'par/par_water_ions.prm']
smol.insertion[smol.resid == 42] = 'A' # Adding an insertion to test that disulfide bonds with insertions work
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir)
compareDir = home(dataDir=os.path.join('test-charmm-build', '3PTB_insertion'))
assertSameAsReferenceDir(compareDir, tmpdir)
def __init__(self, arr, vecMin, vecRes, color=8, isovalue=0.5):
""" Displays an isosurface in VMD
The function returns an instance of VMDGraphicsObject. To delete it, use the delete() method.
Parameters
----------
arr: np.ndarray
3D array with volumetric data.
vecMin: np.ndarray
3D vector denoting the minimal corner of the grid
vecRes: np.ndarray
3D vector denoting the resolution of the grid in each dimension
color: str
The color to be used for the isosurface
"""
super().__init__(arr)
from htmd.util import tempname
import numpy as np
import os
filename = tempname(suffix='.cube')
outFile = open(filename, 'w')
# conversion to gaussian units
L = 1 / 0.52917725
gauss_bin = vecRes * L
# minCorner = 0.5*L*(vecMin - vecMax + vecRes)
minCorner = L * (vecMin + 0.5 * vecRes)
ngrid = arr.shape
# write header
outFile.write("CUBE FILE\n")
outFile.write("OUTER LOOP: X, MIDDLE LOOP: Y, INNER LOOP: Z\n")
outFile.write("%5d %12.6f %12.6f %12.6f\n" % (1, minCorner[0], minCorner[1], minCorner[2]))
outFile.write("%5d %12.6f %12.6f %12.6f\n" % (ngrid[0], gauss_bin[0], 0, 0))
outFile.write("%5d %12.6f %12.6f %12.6f\n" % (ngrid[1], 0, gauss_bin[1], 0))
outFile.write("%5d %12.6f %12.6f %12.6f\n" % (ngrid[2], 0, 0, gauss_bin[2]))
outFile.write("%5d %12.6f %12.6f %12.6f %12.6f\n" % (1, 0, minCorner[0], minCorner[1], minCorner[2]))
# main loop
cont = 0
for i in range(ngrid[0]):
for j in range(ngrid[1]):
for k in range(ngrid[2]):
outFile.write("%13.5g" % arr[i][j][k])
if np.mod(cont, 6) == 5:
outFile.write("\n")
cont += 1
outFile.close()
vmd = getCurrentViewer()
vmd.send('mol new {} type cube first 0 last -1 step 1 waitfor 1 volsets {{0 }}'.format(filename))
vmd.send('mol modstyle top top Isosurface {} 0 2 0 2 1'.format(isovalue))
vmd.send('mol modcolor top top ColorID {}'.format(color))
vmd.send('set htmd_graphics_mol({:d}) [molinfo top]'.format(self.n))
os.unlink(filename)
def test_acemd3(self):
from htmd.util import tempname
from htmd.home import home
from htmd.units import convert
from glob import glob
import os
from htmd.mdengine.acemd.acemd import GroupRestraint
eq = Equilibration()
eq.runtime = 4
eq.timeunits = "ns"
eq.temperature = 300
eq.restraintsteps = convert(
eq.timeunits, "timesteps", eq.runtime, timestep=eq.acemd.timestep
)
ligres = GroupRestraint(
"resname MOL and noh",
[42, 38, 1],
[(5, 0)],
fbcentre=[-0.178, -0.178, 29.195],
)
eq.restraints = eq.defaultEquilRestraints(1000000) + [ligres]
tmpdir = tempname()
eq.write(
home(dataDir=os.path.join("test-protocols", "build", "protLig")), tmpdir
)
# Compare with reference
refdir = home(
dataDir=os.path.join(
"test-protocols", "equilibration", "acemd3", "protLig", "prerun"
)
)
files = [os.path.basename(f) for f in glob(os.path.join(refdir, "*"))]
self._compareResultFolders(refdir, tmpdir, "protLig")
def testWriteDataFile(self):
"""Test case - write data file
"""
self.lfh.write("\nStarting %s %s\n" % (self.__class__.__name__,
sys._getframe().f_code.co_name))
try:
#
myDataList=[]
ofh = open(tempname(suffix='.cif'), "w")
curContainer=DataContainer("myblock")
aCat=DataCategory("pdbx_seqtool_mapping_ref")
aCat.appendAttribute("ordinal")
aCat.appendAttribute("entity_id")
aCat.appendAttribute("auth_mon_id")
aCat.appendAttribute("auth_mon_num")
aCat.appendAttribute("pdb_chain_id")
aCat.appendAttribute("ref_mon_id")
aCat.appendAttribute("ref_mon_num")
aCat.append((1,2,3,4,5,6,7))
aCat.append((1,2,3,4,5,6,7))
aCat.append((1,2,3,4,5,6,7))
aCat.append((1,2,3,4,5,6,7))
curContainer.append(aCat)
myDataList.append(curContainer)
pdbxW=PdbxWriter(ofh)
pdbxW.write(myDataList)
ofh.close()
except:
traceback.print_exc(file=sys.stderr)
self.fail()
def test_production(self):
from htmd.home import home
import filecmp
from htmd.util import tempname
from glob import glob
tmpdir = tempname()
pdbid = '3PTB'
prod = Acemd('production')
prod.run = '2000'
prod.trajectoryfreq = 200
prod.temperature = 300
prod.write(
home(dataDir=os.path.join('test-acemd', pdbid, 'equil_out')),
tmpdir)
print(tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join('test-acemd', pdbid, 'prod'))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
match, mismatch, error = filecmp.cmpfiles(refdir,
tmpdir,
files,
shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError(
'Different results produced by Acemd production for '
'test {} between {} and {} in files {}.'.format(
pdbid, refdir, tmpdir, mismatch))
def test_equilibration(self):
from htmd.home import home
import filecmp
from htmd.util import tempname
from glob import glob
from moleculekit.molecule import Molecule
tmpdir = tempname()
pdbid = "3PTB"
builddir = home(dataDir=os.path.join("test-acemd", pdbid, "build"))
equil = Acemd("equilibration")
mol = Molecule(os.path.join(builddir, "structure.pdb"))
celldim = mol.coords.max(axis=0) - mol.coords.min(axis=0)
equil.boxsize = " ".join(
["{:3.1f}".format(val) for val in celldim.squeeze()])
equil.run = "2000"
equil.trajectoryperiod = 200
equil.temperature = 300
equil.write(builddir, tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join("test-acemd", pdbid, "equil"))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, "*"))]
match, mismatch, error = filecmp.cmpfiles(refdir,
tmpdir,
files,
shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError(
"Different results produced by Acemd equilibration for "
"test {} between {} and {} in files {}.".format(
pdbid, refdir, tmpdir, mismatch))
def test_equilibration(self):
from htmd.home import home
import filecmp
from htmd.util import tempname
from glob import glob
from htmd.molecule.molecule import Molecule
tmpdir = tempname()
pdbid = '3PTB'
builddir = home(dataDir=os.path.join('test-acemd', pdbid, 'build'))
equil = Acemd('equilibration')
mol = Molecule(os.path.join(builddir, 'structure.pdb'))
celldim = mol.coords.max(axis=0) - mol.coords.min(axis=0)
equil.celldimension = ' '.join(
['{:3.1f}'.format(val) for val in celldim.squeeze()])
equil.run = '2000'
equil.trajectoryfreq = 200
equil.temperature = 300
equil.write(builddir, tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join('test-acemd', pdbid, 'equil'))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
match, mismatch, error = filecmp.cmpfiles(refdir,
tmpdir,
files,
shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError(
'Different results produced by Acemd equilibration for '
'test {} between {} and {} in files {}.'.format(
pdbid, refdir, tmpdir, mismatch))
def test_production(self):
from htmd.home import home
import filecmp
from htmd.util import tempname
from glob import glob
tmpdir = tempname()
pdbid = "3PTB"
prod = Acemd("production")
prod.run = "2000"
prod.trajectoryperiod = 200
prod.temperature = 300
prod.write(
home(dataDir=os.path.join("test-acemd", pdbid, "equil_out")),
tmpdir)
print(tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join("test-acemd", pdbid, "prod"))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, "*"))]
match, mismatch, error = filecmp.cmpfiles(refdir,
tmpdir,
files,
shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError(
"Different results produced by Acemd production for "
"test {} between {} and {} in files {}.".format(
pdbid, refdir, tmpdir, mismatch))
def test_h2o2_dihed_opt_restart(self):
refDir = os.path.join(self.dataDir, 'h2o2_dihed_opt_restart')
resDir = tempname()
shutil.copytree(os.path.join(refDir, 'dihedral-opt'),
os.path.join(resDir, 'dihedral-opt'))
self._test(refDir, resDir,
'parameterize -m input.mol2 -f GAFF2 --no-min --no-esp')
def test_h2o2_esp_restart(self):
refDir = os.path.join(self.dataDir, 'h2o2_esp_restart')
resDir = tempname()
shutil.copytree(os.path.join(refDir, 'esp'),
os.path.join(resDir, 'esp'))
self._test(
refDir, resDir,
'parameterize -m input.mol2 -f GAFF2 --no-min --no-torsions')
def setUpClass(self):
from htmd.util import tempname
from htmd.home import home
from moleculekit.projections.metricdistance import MetricDistance
import shutil
import os
tmpdir = tempname()
shutil.copytree(home(dataDir='adaptive'), tmpdir)
os.chdir(tmpdir)
def test_glycol_dihed_fix_restart_2(self):
refDir = os.path.join(self.dataDir, 'glycol_dihed_fix')
resDir = tempname()
shutil.copytree(os.path.join(refDir, 'dihedral-single-point'),
os.path.join(resDir, 'dihedral-single-point'))
self._test(
refDir, resDir,
'parameterize input.mol2 -d O1-C1-C2-O2 C1-C2-O2-H6 --no-min --no-esp --no-dihed-opt'
)
def test_benzamidine_rtf_prm(self):
refDir = os.path.join(self.dataDir, 'benzamidine_rtf_prm')
resDir = tempname()
os.makedirs(resDir)
shutil.copy(os.path.join(refDir, 'input.rtf'), resDir)
shutil.copy(os.path.join(refDir, 'input.prm'), resDir)
self._test(
refDir, resDir,
'parameterize -m input.mol2 --charge 1 --forcefield CGENFF --rtf input.rtf --prm input.prm --no-min '
'--no-esp --no-torsions')
def setUp(self):
if os.environ.get('TRAVIS_OS_NAME') == 'osx':
self.skipTest('Mac does not work!')
self.maxDiff = None
self.dataDir = home(dataDir='test-param')
if os.environ.get('TESTDIR') is not None:
self.testDir = os.environ['TESTDIR']
else:
self.testDir = tempname()
print('Test dir: {}'.format(self.testDir))
def test_benzamidine_esp_freeze_restart(self):
refDir = os.path.join(self.dataDir, 'benzamidine_esp_freeze_restart')
resDir = tempname()
shutil.copytree(os.path.join(refDir, 'minimize'),
os.path.join(resDir, 'minimize'))
shutil.copytree(os.path.join(refDir, 'esp'),
os.path.join(resDir, 'esp'))
self._test(
refDir, resDir,
'parameterize -m input.mol2 --charge 1 --basis 6-31g-star --no-torsions --freeze-charge N2'
)
def fixParameters(parameterfile):
from htmd.util import tempname
tmpfile = tempname(suffix='.prm')
with open(parameterfile, 'r') as f:
lines = f.readlines()
with open(tmpfile, 'w') as f:
for l in lines:
l = l.replace('!MASS', 'MASS')
l = l.replace('!ATOMS', 'ATOMS')
f.write(l)
return tmpfile
def test_benzamidine_full_restart(self):
refDir = os.path.join(self.dataDir, 'benzamidine_full_restart')
resDir = tempname()
shutil.copytree(os.path.join(refDir, 'minimize'),
os.path.join(resDir, 'minimize'))
shutil.copytree(os.path.join(refDir, 'esp'),
os.path.join(resDir, 'esp'))
shutil.copytree(os.path.join(refDir, 'dihedral-opt'),
os.path.join(resDir, 'dihedral-opt'))
self._test(refDir, resDir,
'parameterize -m input.mol2 --charge 1 --basis 6-31g-star')
def fixParameters(parameterfile):
from htmd.util import tempname
tmpfile = tempname(suffix='.prm')
with open(parameterfile, 'r') as f:
lines = f.readlines()
with open(tmpfile, 'w') as f:
for l in lines:
l = l.replace('!MASS', 'MASS')
l = l.replace('!ATOMS', 'ATOMS')
f.write(l)
return tmpfile
def testSimpleInitialization(self):
"""Test case - Simple initialization of a data category and data block
"""
self.lfh.write("\nStarting %s %s\n" % (self.__class__.__name__,
sys._getframe().f_code.co_name))
try:
#
fn = tempname(suffix='.cif')
attributeNameList=['aOne','aTwo','aThree','aFour','aFive','aSix','aSeven','aEight','aNine','aTen']
rowList=[[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10],
[1,2,3,4,5,6,7,8,9,10]
]
nameCat='myCategory'
#
#
curContainer=DataContainer("myblock")
aCat=DataCategory(nameCat,attributeNameList,rowList)
aCat.printIt()
curContainer.append(aCat)
curContainer.printIt()
#
myContainerList=[]
myContainerList.append(curContainer)
ofh = open(fn, "w")
pdbxW=PdbxWriter(ofh)
pdbxW.write(myContainerList)
ofh.close()
myContainerList=[]
ifh = open(fn, "r")
pRd=PdbxReader(ifh)
pRd.read(myContainerList)
ifh.close()
for container in myContainerList:
for objName in container.getObjNameList():
name,aList,rList=container.getObj(objName).get()
self.lfh.write("Recovered data category %s\n" % name)
self.lfh.write("Attribute list %r\n" % repr(aList))
self.lfh.write("Row list %r\n" % repr(rList))
except:
traceback.print_exc(file=self.lfh)
self.fail()
def test_customDisulfideBonds(self):
from htmd.molecule.molecule import Molecule
from htmd.builder.solvate import solvate
from htmd.home import home
from htmd.util import tempname, assertSameAsReferenceDir
import os
import numpy as np
# Use pre-prepared files so we can tell whether the error is in prepare or in build
# Inputs are reference outputs of proteinprepare.
preparedInputDir = home(dataDir='test-proteinprepare')
pdb = '1GZM'
inFile = os.path.join(preparedInputDir, pdb, "{}-prepared.pdb".format(pdb))
mol = Molecule(inFile)
mol.filter('protein') # Fix for bad proteinPrepare hydrogen placing
np.random.seed(1) # Needed for ions
smol = solvate(mol)
topos = ['top/top_all36_prot.rtf', 'top/top_water_ions.rtf']
params = ['par/par_all36_prot_mod.prm', 'par/par_water_ions.prm']
disu = [['segid 1 and resid 110', 'segid 1 and resid 187'], ['segid 0 and resid 110', 'segid 0 and resid 187']]
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir, disulfide=disu)
compareDir = home(dataDir=os.path.join('test-charmm-build', pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
# TODO: Remove this after deprecation
from htmd.builder.builder import DisulfideBridge
np.random.seed(1) # Needed for ions
disu = [DisulfideBridge('1', 110, '1', 187), DisulfideBridge('0', 110, '0', 187)]
tmpdir = tempname()
_ = build(smol, topo=topos, param=params, outdir=tmpdir, disulfide=disu)
compareDir = home(dataDir=os.path.join('test-charmm-build', pdb))
assertSameAsReferenceDir(compareDir, tmpdir)
def test_production(self):
from htmd.home import home
import filecmp
from htmd.util import tempname
from glob import glob
tmpdir = tempname()
pdbid = '3PTB'
prod = Acemd('production')
prod.run = '2000'
prod.trajectoryfreq = 200
prod.temperature = 300
prod.write(home(dataDir=os.path.join('test-acemd', pdbid, 'equil_out')), tmpdir)
print(tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join('test-acemd', pdbid, 'prod'))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
match, mismatch, error = filecmp.cmpfiles(refdir, tmpdir, files, shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError('Different results produced by Acemd production for '
'test {} between {} and {} in files {}.'.format(pdbid, refdir, tmpdir, mismatch))
def test_glycol_dihed_fix_restart_2(self):
refDir = os.path.join(self.dataDir, 'glycol_dihed_fix')
resDir = tempname()
shutil.copytree(os.path.join(refDir, 'dihedral-single-point'), os.path.join(resDir, 'dihedral-single-point'))
self._test(refDir, resDir, 'parameterize input.mol2 -d O1-C1-C2-O2 C1-C2-O2-H6 --no-min --no-esp --no-dihed-opt')
for i in range(len(St)):
mst[i] = macro_ofcluster[St[i]]
return mst
'''def _macroP(C, macro_ofmicro):
macronum = np.max(macro_ofmicro) + 1
macroC = np.zeros((macronum, macronum))
for m1 in range(macronum):
sourcemicros = np.where(macro_ofmicro == m1)[0]
for m2 in range(macronum):
sinkmicros = np.where(macro_ofmicro == m2)[0]
ixgrid = np.ix_(sourcemicros, sinkmicros)
macroC[m1, m2] = np.sum(C[ixgrid].flatten())
from msmtools.estimation import transition_matrix
return transition_matrix(macroC, reversible=True)'''
if __name__ == '__main__':
from htmd.util import tempname
from htmd.home import home
from os.path import join
testfolder = home(dataDir='model')
model = Model(file=join(testfolder, 'model.dat'))
tmpsave = tempname(suffix='.dat')
model.save(tmpsave)
def _prepareStream(filename):
from htmd.util import tempname
tmpout = tempname()
os.makedirs(tmpout)
return split(filename, tmpout)
def setUp(self):
from htmd.util import tempname
self.testDir = os.environ.get('TESTDIR', tempname())
print('Running tests in {}'.format(self.testDir))
numsteps = convert(self.timeunits, 'timesteps', self.runtime, timestep=self.acemd.timestep)
self.acemd.temperature = str(self.temperature)
self.acemd.langevintemp = str(self.temperature)
if self.fb_k > 0: #use TCL only for flatbottom
mol = Molecule(os.path.join(inputdir, self.acemd.coordinates))
self.acemd.tclforces = 'on'
tcl = list(self.acemd.TCL)
tcl[0] = tcl[0].format(NUMSTEPS=numsteps, KCONST=self.fb_k,
REFINDEX=' '.join(map(str, mol.get('index', self.fb_reference))),
SELINDEX=' '.join(map(str, mol.get('index', self.fb_selection))),
BOX=' '.join(map(str, self.fb_box)))
self.acemd.TCL = tcl[0] + tcl[1]
else:
self.acemd.TCL = 'set numsteps {}\n'.format(numsteps)
self.acemd.setup(inputdir, outputdir, overwrite=True)
if __name__ == "__main__":
import htmd.home
from htmd.util import tempname
md = Production()
md.temperature = 300
md.fb_reference = 'protein and name CA'
md.fb_selection = 'segname L and noh'
md.acemd.extendedsystem = None # use different data
md.acemd.binindex = None # use different data
md.fb_box = [-20, 20, -20, 20, 43, 45]
md.fb_k = 5
md.write(htmd.home.home() +'/data/equilibrate', tempname())
md.fb_k = 0
md.write(htmd.home.home() +'/data/equilibrate', tempname())
def dock(protein, ligand, center=None, extent=None, numposes=20, babelexe='babel', vinaexe=None):
""" Molecular docking, using Vina
If centre and extent are not provided, docking will be performed over the whole protein
Parameters
----------
protein : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
Molecule object representing the receptor
ligand : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
Molecule object representing the ligand to dock
center : list
3-vec centre of of the search bounding box (optional)
extent : list
3-vec linear extent of the search bounding box (optional)
numposes : int
Number of poses to return. Vina cannot return more than 20 poses.
babelexe : str
Path to babel executable.
vinaexe : str
Path to AutoDock Vina executable.
Returns
-------
poses
Molecule object representing the N<10 best poses
scores
3x num_poses matrix containing kcal, rmsd lb, rmsd ub
Examples
--------
>>> poses, scoring = dock(protein, ligand)
>>> poses, scoring = dock(protein, ligand, center=[ 10., 5., 12. ], extent=[ 15., 15., 15. ] )
"""
if np.size(protein.coords, 2) != 1 or np.size(ligand.coords, 2) != 1:
raise NameError('Protein and ligand Molecules should be single frames')
buffer = 10. # Angstrom buffer around protein for whole-protein docking
c_min = np.min(protein.coords, 0).reshape((1, 3))[0]
c_max = np.max(protein.coords, 0).reshape((1, 3))[0]
if center is None:
center = (buffer + (c_max + c_min)) / 2
if extent is None:
extent = (c_max - c_min) + buffer
# babel -i pdb protein.pdb -o pdbqt protein.pdbqt -xr
# babel -i pdb ligand.pdb -o pdbqt ligand.pdbqt -xhn
# vina --ligand ligand.pdbqt --receptor protein.pdbqt --center_x 0. --center_y 0. --center_z 0. --size_x 60. --size_y 60. --size_z 60 --exhaustiveness 10
# babel -m -i pdbqt ligand_out.pdbqt -o pdb out_.pdb -xhn
protein_pdb = tempname(suffix=".pdb")
ligand_pdb = tempname(suffix=".pdb")
output_pdb = tempname(suffix="_.pdb")
output_prefix = path.splitext(output_pdb)[0]
protein_pdbqt = tempname(suffix=".pdbqt")
ligand_pdbqt = tempname(suffix=".pdbqt")
output_pdbqt = tempname(suffix=".pdbqt")
protein.write(protein_pdb)
ligand.write(ligand_pdb)
try:
if vinaexe is None:
import platform
suffix = ''
if platform.system() == "Windows":
suffix = '.exe'
vinaexe = '{}-vina{}'.format(platform.system(), suffix)
vinaexe = shutil.which(vinaexe, mode=os.X_OK)
if not vinaexe:
raise NameError('Could not find vina, or no execute permissions are given')
except:
raise NameError('Could not find vina, or no execute permissions are given')
try:
babelexe = shutil.which(babelexe, mode=os.X_OK)
if babelexe is None:
raise NameError('Could not find babel, or no execute permissions are given')
except:
raise NameError('Could not find babel, or no execute permissions are given')
call([babelexe, '-i', 'pdb', protein_pdb, '-o', 'pdbqt', protein_pdbqt, '-xr'])
call([babelexe, '-i', 'pdb', ligand_pdb, '-o', 'pdbqt', ligand_pdbqt, '-xhn'])
if not path.isfile(ligand_pdbqt):
raise NameError('Ligand could not be converted to PDBQT')
if not path.isfile(protein_pdbqt):
raise NameError('Protein could not be converted to PDBQT')
call([vinaexe, '--receptor', protein_pdbqt, '--ligand', ligand_pdbqt, '--out', output_pdbqt,
'--center_x', str(center[0]), '--center_y', str(center[1]), '--center_z', str(center[2]),
'--size_x', str(extent[0]), '--size_y', str(extent[1]), '--size_z', str(extent[2]), '--num_modes', str(numposes)])
call([babelexe, '-m', '-i', 'pdbqt', output_pdbqt, '-o', 'pdb', output_pdb, '-xhn'])
scoring = []
coords = []
outfiles = natsorted(glob('{}*.pdb'.format(output_prefix)))
for outf in outfiles:
# First get the scoring
scoring.append(_parseScoring(outf))
next_pose = Molecule(outf)
os.remove(outf)
c = next_pose.coords
co = c.copy()
natoms = len(ligand.name)
# Order atoms back to original order
for idx_i in range(natoms):
for idx_j in range(natoms):
if ligand.name[idx_i] == next_pose.name[idx_j]:
co[idx_i, :, :] = c[idx_j, :, :]
coords.append(co)
poses = []
for i, c in enumerate(coords):
l = ligand.copy()
l.viewname = 'Pose {}'.format(i)
l.coords = c
poses.append(l)
os.remove(protein_pdb)
os.remove(ligand_pdb)
os.remove(protein_pdbqt)
os.remove(ligand_pdbqt)
os.remove(output_pdbqt)
return poses, np.array(scoring)
def loopModeller(
mol,
segid,
seq,
startresid,
movstart=None,
movend=None,
modellerexe="/shared/sdoerr/Software/modeller-9.16/build/bin/mod9.16",
):
""" Uses the Modeller software to predict missing loops in a Molecule.
Parameters
----------
mol : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
A Molecule object.
segid : str
The name of the segment containing the gap.
seq : str
The sequence of residues to be added by the loop modeller.
startresid : int
The resid of the residue before the gap.
movstart : int
The resid after which the residues will be allowed to move in modeller.
movend : int
The resid before which the residues will be allowed to move in modeller.
modellerexe : str
The path to the modeller executable.
Returns
-------
newmol : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
A new Molecule object containing the protein with the modelled loop.
Examples
--------
>>> newmol = loopModeller(mol, 'B', 'MLSGSK', 263)
"""
if movstart is None:
movstart = startresid
if movend is None:
movend = movstart + len(seq) + 1
segatm = mol.atomselect("segid {}".format(segid))
segres = np.unique(mol.get("resid", sel=segatm))
chain = np.unique(mol.get("chain", sel=segatm))
if len(chain) != 1:
raise RuntimeError("More than one chain detected in segment {}".format(segid))
pos = np.where(segres == startresid)[0][0]
if (segres[pos + 1] - segres[pos]) != (len(seq) + 1):
raise RuntimeError(
"Sequence is {} characters long while sequence gap ({}-{}) is {} long. Cannot insert loop.".format(
len(seq), segres[pos], segres[pos + 1], (segres[pos + 1] - segres[pos])
)
)
chain = chain[0]
segresidstart = np.min(segres)
segresidend = np.max(segres)
currseq = mol.sequence()[segid]
minuses = ["-"] * len(seq)
minuses = "".join(minuses)
inspos = np.where(segres == startresid)[0] + 1
gapseq = currseq[:inspos] + minuses + currseq[inspos:] + "*"
fullseq = currseq[:inspos] + seq + currseq[inspos:] + "*"
# Get the sequence of the 1qg8 PDB file, and write to an alignment file
pdbfile = tempname(suffix=".pdb")
mol.write(pdbfile)
alifile = tempname(suffix=".ali")
f = open(alifile, "w")
f.write(">P1;prot\nstructure:{}:{}:{}:{}:{}::::\n".format(pdbfile, segresidstart, chain, segresidend, chain))
f.write(gapseq)
f.write("\n>P1;prot_fill\nsequence:::::::::\n")
f.write(fullseq)
f.close()
script = """
from modeller import *
from modeller.automodel import * # Load the automodel class
log.verbose()
env = environ()
class MyModel(automodel):
def select_atoms(self):
return selection(self.residue_range('{}', '{}'))
a = MyModel(env, alnfile = '{}', knowns = 'prot', sequence = 'prot_fill')
a.starting_model= 1
a.ending_model = 1
a.make()
""".format(
movstart + 1 - segresidstart, movend - segresidstart, alifile
)
pyfile = tempname(suffix=".py")
f = open(pyfile, "w")
f.write(script)
f.close()
print(pdbfile)
print(alifile)
print(pyfile)
call("{} {}".format(modellerexe, pyfile), shell=True)
newmol = Molecule("./prot_fill.B99990001.pdb")
return newmol
bind = {}
for i, elt in enumerate(list(set(b))):
bind[elt] = i
return np.array([bind.get(itm, -1) for itm in a]) # None can be replaced by any other "not in b" value
if __name__ == '__main__':
from htmd import *
from htmd.util import tempname
from htmd.home import home
from os.path import join
testfolder = home(dataDir='adaptive')
sims = simlist(glob(join(testfolder, 'data', '*', '')), glob(join(testfolder, 'input', '*', 'structure.pdb')))
fsims = simfilter(sims, tempname(), 'not water')
metr = Metric(fsims)
metr.set(MetricDistance('protein and resid 10 and name CA', 'resname BEN and noh', metric='contacts',
groupsel1='residue', threshold=4))
data1 = metr.project()
metr.set(MetricDihedral())
data2 = metr.project()
# Testing combining of metrics
data1.combine(data2)
# Testing dimensions
assert np.array_equal(data1.map.shape, (897, 3)), 'combine not working correct'
assert np.array_equal(data1.dat[0].shape, (6, 897)), 'combine not working correct'
assert np.array_equal(np.where(data1.map.type == 'contact')[0], [0, 1, 2, 3, 4, 5, 6, 7, 8]), 'combine not working correct'
def dock(protein, ligand, center=None, extent=None, numposes=20, babelexe='obabel', vinaexe=None):
""" Molecular docking, using Vina
If centre and extent are not provided, docking will be performed over the whole protein
Parameters
----------
protein : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
Molecule object representing the receptor
ligand : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
Molecule object representing the ligand to dock
center : list
3-vec centre of of the search bounding box (optional)
extent : list
3-vec linear extent of the search bounding box (optional)
numposes : int
Number of poses to return. Vina cannot return more than 20 poses.
babelexe : str
Path to babel executable.
vinaexe : str
Path to AutoDock Vina executable.
Returns
-------
poses
Molecule object representing the N<10 best poses
scores
3x num_poses matrix containing kcal, rmsd lb, rmsd ub
Examples
--------
>>> poses, scoring = dock(protein, ligand)
>>> poses, scoring = dock(protein, ligand, center=[ 10., 5., 12. ], extent=[ 15., 15., 15. ] )
"""
if np.size(protein.coords, 2) != 1 or np.size(ligand.coords, 2) != 1:
raise NameError('Protein and ligand Molecules should be single frames')
buffer = 10. # Angstrom buffer around protein for whole-protein docking
c_min = np.min(protein.coords, 0).reshape((1, 3))[0]
c_max = np.max(protein.coords, 0).reshape((1, 3))[0]
if center is None:
center = (buffer + (c_max + c_min)) / 2
if extent is None:
extent = (c_max - c_min) + buffer
# babel -i pdb protein.pdb -o pdbqt protein.pdbqt -xr
# babel -i pdb ligand.pdb -o pdbqt ligand.pdbqt -xhn
# vina --ligand ligand.pdbqt --receptor protein.pdbqt --center_x 0. --center_y 0. --center_z 0. --size_x 60. --size_y 60. --size_z 60 --exhaustiveness 10
# babel -m -i pdbqt ligand_out.pdbqt -o pdb out_.pdb -xhn
protein_pdb = tempname(suffix=".pdb")
ligand_mol2 = tempname(suffix=".mol2")
output_pdb = tempname(suffix="_.pdb")
output_prefix = path.splitext(output_pdb)[0]
protein_pdbqt = tempname(suffix=".pdbqt")
ligand_pdbqt = tempname(suffix=".pdbqt")
output_pdbqt = tempname(suffix=".pdbqt")
protein.write(protein_pdb)
lig2 = ligand.copy()
lig2.atomtype = lig2.element # babel does not understand mol2 atomtypes and requires elements instead
lig2.write(ligand_mol2)
# Dirty hack to remove the 'END' line from the PDBs since babel hates it
with open(protein_pdb, 'r') as f:
lines = f.readlines()
with open(protein_pdb, 'w') as f:
f.writelines(lines[:-1])
# End of dirty hack
try:
if vinaexe is None:
import platform
suffix = ''
if platform.system() == "Windows":
suffix = '.exe'
vinaexe = '{}-vina{}'.format(platform.system(), suffix)
vinaexe = shutil.which(vinaexe, mode=os.X_OK)
if not vinaexe:
raise NameError('Could not find vina, or no execute permissions are given')
except:
raise NameError('Could not find vina, or no execute permissions are given')
try:
babelexe = shutil.which(babelexe, mode=os.X_OK)
if babelexe is None:
raise NameError('Could not find babel, or no execute permissions are given')
except:
raise NameError('Could not find babel, or no execute permissions are given')
call([babelexe, '-i', 'pdb', protein_pdb, '-o', 'pdbqt', '-O', protein_pdbqt, '-xr'])
if np.all(ligand.charge != 0):
logger.info('Charges detected in ligand and will be used for docking.')
call([babelexe, '-i', 'mol2', ligand_mol2, '-o', 'pdbqt', '-O', ligand_pdbqt, '-xn', '-xh'])
else:
logger.info('Charges were not defined for all atoms. Will guess charges anew using gasteiger method.')
call([babelexe, '-i', 'mol2', ligand_mol2, '-o', 'pdbqt', '-O', ligand_pdbqt, '-xn', '-xh', '--partialcharge', 'gasteiger'])
if not path.isfile(ligand_pdbqt):
raise NameError('Ligand could not be converted to PDBQT')
if not path.isfile(protein_pdbqt):
raise NameError('Protein could not be converted to PDBQT')
call([vinaexe, '--receptor', protein_pdbqt, '--ligand', ligand_pdbqt, '--out', output_pdbqt,
'--center_x', str(center[0]), '--center_y', str(center[1]), '--center_z', str(center[2]),
'--size_x', str(extent[0]), '--size_y', str(extent[1]), '--size_z', str(extent[2]), '--num_modes', str(numposes)])
call([babelexe, '-m', '-i', 'pdbqt', output_pdbqt, '-o', 'pdb', '-O', output_pdb, '-xhn'])
from natsort import natsorted
outfiles = natsorted(glob('{}*.pdb'.format(output_prefix)))
scoring = []
poses = []
for i, ligf in enumerate(outfiles):
scoring.append(_parseScoring(ligf))
l = Molecule(ligf)
l.viewname = 'Pose {}'.format(i)
poses.append(l)
os.remove(protein_pdb)
os.remove(ligand_mol2)
os.remove(protein_pdbqt)
os.remove(ligand_pdbqt)
os.remove(output_pdbqt)
return poses, np.array(scoring)
def setUp(self):
self.lfh=sys.stderr
self.verbose=False
self.pathPdbxDataFile = os.path.join(home(dataDir='molecule-readers'), "1kip.cif")
self.pathOutputFile = tempname(suffix='.cif')
from htmd.home import home
from htmd.molecule.molecule import Molecule, mol_equal
from htmd.util import tempname
import numpy as np
import os
testfolder = home(dataDir='metricdistance')
mol = Molecule(os.path.join(testfolder, 'filtered.pdb'))
mol.coords = np.tile(mol.coords, (1, 1, 2))
mol.filter('protein and resid 1 to 20')
mol.boxangles = np.ones((3, 2), dtype=np.float32) * 90
mol.box = np.ones((3, 2), dtype=np.float32) * 15
mol.step = np.arange(2)
mol.time = np.arange(2) * 1E5
for ext in _WRITERS:
tmp = tempname(suffix='.'+ext)
mol.write(tmp)
print('Can write {} files'.format(ext))
# from difflib import Differ
# d = Differ()
#
# with open(tmp, 'rb') as f1, open(a, 'rb') as f2:
# res = d.compare([x.decode('utf8') for x in f1.readlines()], [x.decode('utf8') for x in f2.readlines()])
#
# try:
# next(res)
# except StopIteration:
# print('Same files')
# else:
def dock(protein, ligand, center=None, extent=None):
""" Molecular docking, using Vina
If centre and extent are not provided, docking will be performed over the whole protein
Parameters
----------
protein
Molecule object representing the receptor
ligand
Molecule object representing the ligand to dock
center
3-vec centre of of the search bounding box (optional)
extent
3-vec linear extent of the search bounding box (optional)
Returns
-------
poses
Molecule object representing the N<10 best poses
scores
3x num_poses matrix containing kcal, rmsd lb, rmsd ub
Examples
--------
>>> poses = dock(protein, ligand)
>>> poses = dock(protein, ligand, center=[ 10., 5., 12. ], extent=[ 15., 15., 15. ] )
"""
if np.size(protein.coords, 2) != 1 or np.size(ligand.coords, 2) != 1:
raise NameError('Protein and ligand Molecules should be single frames')
buffer = 10. # Angstrom buffer around protein for whole-protein docking
c_min = np.min(protein.coords, 0).reshape((1, 3))[0]
c_max = np.max(protein.coords, 0).reshape((1, 3))[0]
if center is None:
center = (buffer + (c_max + c_min)) / 2
if extent is None:
extent = (c_max - c_min) + buffer
# babel -i pdb protein.pdb -o pdbqt protein.pdbqt -xr
# babel -i pdb ligand.pdb -o pdbqt ligand.pdbqt -xhn
# vina --ligand ligand.pdbqt --receptor protein.pdbqt --center_x 0. --center_y 0. --center_z 0. --size_x 60. --size_y 60. --size_z 60 --exhaustiveness 10
# babel -m -i pdbqt ligand_out.pdbqt -o pdb out_.pdb -xhn
protein_pdb = tempname(suffix=".pdb")
ligand_pdb = tempname(suffix=".pdb")
output_pdb = tempname(suffix="_.pdb")
output_prefix = path.splitext(output_pdb)[0]
protein_pdbqt = tempname(suffix=".pdbqt")
ligand_pdbqt = tempname(suffix=".pdbqt")
output_pdbqt = tempname(suffix=".pdbqt")
protein.write(protein_pdb)
ligand.write(ligand_pdb)
try:
import platform
suffix=""
if platform.system() == "Windows":
suffix=".exe"
vinaexe = shutil.which(platform.system() + "-vina" + suffix, mode=os.X_OK )
if not vinaexe:
raise NameError('Could not find vina, or no execute permissions are given')
except:
raise NameError('Could not find vina, or no execute permissions are given')
try:
babelexe = shutil.which('htmd_babel', mode=os.X_OK)
except:
raise NameError('Could not find babel, or no execute permissions are given')
#babelexe = path.join(home(), 'bin', 'htmd_babel')
#vinaexe = path.join(home(), 'bin', 'htmd_vina')
#if not os.access(babelexe, os.X_OK):
# raise NameError('Could not find ' + babelexe + ' or no execute permissions are given')
#if not os.access(vinaexe, os.X_OK):
# raise NameError('Could not find ' + vinaexe + ' or no execute permissions are given')
#from IPython.core.debugger import Tracer
#Tracer()()
call([babelexe, '-i', 'pdb', protein_pdb, '-o', 'pdbqt', protein_pdbqt, '-xr'])
call([babelexe, '-i', 'pdb', ligand_pdb, '-o', 'pdbqt', ligand_pdbqt, '-xhn'])
if not path.isfile(ligand_pdbqt):
raise NameError('Ligand could not be converted to PDBQT')
if not path.isfile(protein_pdbqt):
raise NameError('Protein could not be converted to PDBQT')
call([vinaexe, '--receptor', protein_pdbqt, '--ligand', ligand_pdbqt, '--out', output_pdbqt,
'--center_x', str(center[0]), '--center_y', str(center[1]), '--center_z', str(center[2]),
'--size_x', str(extent[0]), '--size_y', str(extent[1]), '--size_z', str(extent[2])])
call([babelexe, '-m', '-i', 'pdbqt', output_pdbqt, '-o', 'pdb', output_pdb, '-xhn'])
scoring = np.zeros((0,3))
coords = []
idx = 1
name = '{}{}.pdb'.format(output_prefix, idx)
while path.isfile(name):
# First get the scoring
kcal = float(check_output('grep "VINA RESULT" ' + name + ' | awk \'{print $4}\'', shell=True).decode('ascii').strip())
rmsdlb = float(check_output('grep "VINA RESULT" ' + name + ' | awk \'{print $5}\'', shell=True).decode('ascii').strip())
rmsdub = float(check_output('grep "VINA RESULT" ' + name + ' | awk \'{print $6}\'', shell=True).decode('ascii').strip())
scoring = np.append(scoring, np.array([[float(kcal), float(rmsdlb), float(rmsdub)]]), axis=0)
next_pose = Molecule(name)
os.remove(name)
c = next_pose.coords
co = c.copy()
natoms = len(ligand.name)
for idx_i in range(natoms):
for idx_j in range(natoms):
if ligand.name[idx_i] == next_pose.name[idx_j]:
co[idx_i, :, :] = c[idx_j, :, :]
'''if idx == 1:
coords = co
else:
coords = np.append(coords, co, axis=2)'''
coords.append(co)
idx += 1
name = '{}{}.pdb'.format(output_prefix, idx)
poses = []
for i, c in enumerate(coords):
l = ligand.copy()
l.viewname = 'Pose {}'.format(i)
l.coords = c
poses.append(l)
os.remove(protein_pdb)
os.remove(ligand_pdb)
os.remove(protein_pdbqt)
os.remove(ligand_pdbqt)
os.remove(output_pdbqt)
return poses, scoring
def test_acemd2(self):
from htmd.home import home
from htmd.util import tempname
import filecmp
from glob import glob
tmpdir = tempname()
acemd = Acemd2()
acemd.temperature = 350
acemd.restart = 'on'
acemd.restartfreq = 25
acemd.outputname = 'output'
acemd.xtcfile = 'myout.xtc'
acemd.xtcfreq = 500
acemd.timestep = 4
acemd.rigidbonds = 'all'
acemd.hydrogenscale = 0.3
acemd.switching = 'on'
acemd.switchdist = 9
acemd.cutoff = 5
acemd.exclude = 'scaled1-4'
acemd.scaling14 = 5.6
acemd.langevin = 'on'
acemd.langevintemp = 300
acemd.langevindamping = 0.8
acemd.pme = 'on'
acemd.pmegridspacing = 5
acemd.fullelectfrequency = 3
acemd.energyfreq = 10
acemd.constraints = 'on'
acemd.consref = '5dhfr_cube.pdb'
acemd.constraintscaling = 3.2
acemd.berendsenpressure = 'on'
acemd.berendsenpressuretarget = 14
acemd.berendsenpressurerelaxationtime = 2
acemd.tclforces = 'on'
acemd.minimize = 150
acemd.run = 555
acemd.celldimension = [3, 56, 2]
acemd.useconstantratio = 'on'
acemd.amber = 'off'
acemd.dielectric = 16
acemd.pairlistdist = 9
acemd.TCL = 'on'
acemd.bincoordinates = None
acemd.binvelocities = None
acemd.binindex = None
acemd.structure = '5dhfr_cube.psf'
acemd.parameters = 'par_all22_prot.inp'
acemd.extendedsystem = None
acemd.coordinates = '5dhfr_cube.pdb'
acemd.velocities = None
acemd.parmfile = None
acemd.coordinates = '5dhfr_cube.pdb'
acemd.setup(home(dataDir='dhfr/'), tmpdir, overwrite=True)
# Compare with reference
refdir = home(dataDir='test-acemd-v2')
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
match, mismatch, error = filecmp.cmpfiles(refdir, tmpdir, files, shallow=False)
if len(mismatch) != 0 or len(error) != 0 or len(match) != len(files):
raise RuntimeError('Different results produced by Acemd.write between {} and {} '
'in files {}.'.format(refdir, tmpdir, mismatch))
import shutil
shutil.rmtree(tmpdir)
for f in deletefiles:
os.remove(f)
# Test ACEMD version 2
pd = Production(_version=2)
pd.runtime = 4
pd.timeunits = 'ns'
pd.temperature = 300
pd.useconstraints = True
pd.constraints = {'protein and name CA': 1, 'protein and noh and not name CA': 0.1}
pd.fb_reference = 'protein and name CA'
pd.fb_selection = 'resname MOL and noh'
pd.fb_box = [-21, 21, -19, 19, 29, 30]
pd.fb_k = 5
tmpdir = tempname()
pd.write(home(dataDir=os.path.join('test-protocols', 'equilibration', 'acemd2', 'protLig', 'postrun')), tmpdir)
# Compare with reference
refdir = home(dataDir=os.path.join('test-protocols', 'production', 'acemd2', 'protLig', 'prerun'))
files = [os.path.basename(f) for f in glob(os.path.join(refdir, '*'))]
_compareResultFolders(refdir, tmpdir, 'protLig')
# Test new ACEMD version 3
pd = Production(_version=3)
pd.runtime = 4
pd.timeunits = 'ns'
pd.temperature = 300
pd.useconstraints = True
pd.constraints = {'protein and name CA': 1, 'protein and noh and not name CA': 0.1}
def buildMembrane(xysize, ratioupper, ratiolower, waterbuff=20, equilibrate=True, outdir='./buildmemb/', lipidf=None):
""" Construct a membrane containing arbitrary lipids and ratios of them.
Parameters
----------
xysize : list
A list containing the size in x and y dimensions of the membrane in Angstroms
ratioupper : list of lists
A list containing sublists indicating the molecule name and the ratio of that molecule for the upper layer
ratiolower : list of lists
Same as ratioupper but for the lower layer
waterbuff : float
The z-dimension size of the water box above and below the membrane
equilibrate : bool
If True it equilibrates the membrane
outdir : str
A folder in which to store the psf and pdb files
lipidf : str
The path to the starting lipid conformations
Returns
-------
mol : :class:`Molecule <htmd.molecule.molecule.Molecule`
The resulting membrane including surrounding waters
Examples
--------
>>> lipidratioupper = [['popc', 10], ['chl1', 1]]
>>> lipidratiolower = [['popc', 8], ['chl1', 2]]
>>> width = [50, 100]
>>> res = buildMembrane(width, lipidratioupper, lipidratiolower)
"""
from htmd.membranebuilder.ringpenetration import resolveRingPenetrations
from htmd.builder.solvate import solvate
from htmd.builder.charmm import build
from htmd.util import tempname
from htmd.molecule.molecule import Molecule
from htmd.home import home
import os
import pandas as pd
if lipidf is None:
lipidf = os.path.join(home(), 'membranebuilder', 'lipids')
lipiddb = pd.read_csv(os.path.join(home(), 'membranebuilder', 'lipiddb.csv'), index_col='Name')
uqlip = np.unique([x[0] for x in ratioupper] + [x[0] for x in ratiolower])
files = _locateLipidFiles(lipidf, uqlip)
area = np.prod(xysize)
lipids = _createLipids(ratioupper, area, lipiddb, files, leaflet='upper')
lipids += _createLipids(ratiolower, area, lipiddb, files, leaflet='lower')
_setPositionsLJSim(xysize, [l for l in lipids if l.xyz[2] > 0])
_setPositionsLJSim(xysize, [l for l in lipids if l.xyz[2] < 0])
_findNeighbours(lipids, xysize)
_loadMolecules(lipids, files)
resolveRingPenetrations(lipids, xysize)
memb = _createMembraneMolecule(lipids)
# from globalminimization import minimize
# newpos, newrot = minimize(lipids, xysize + [100], stepxy=0.5, steprot=50, contactthresh=1)
# for i in range(len(lipids)):
# lipids[i].xyz[:2] = newpos[i]
# lipids[i].rot = newrot[i]
#
# resolveRingPenetrations(lipids, xysize)
# endmemb = _createMembraneMolecule(lipids)
minc = memb.get('coords', 'name P').min(axis=0) - 5
maxc = memb.get('coords', 'name P').max(axis=0) + 5
mm = [[0, 0, maxc[2] - 2], [xysize[0], xysize[1], maxc[2] + waterbuff]]
smemb = solvate(memb, minmax=mm)
mm = [[0, 0, minc[2] - waterbuff], [xysize[0], xysize[1], minc[2] + 2]]
smemb = solvate(smemb, minmax=mm)
smemb.moveBy([0, 0, -smemb.coords[:, 2, 0].min()])
if outdir is None:
outdir = tempname()
print('Outdir ', outdir)
res = build(smemb, ionize=False, stream=['str/lipid/toppar_all36_lipid_cholesterol_model_1.str'], outdir=outdir)
if equilibrate:
from htmd.membranebuilder.simulate_openmm import equilibrateSystem
from shutil import copy, move
outpdb = tempname(suffix='.pdb')
charmmf = os.path.join(home(), 'membranebuilder', 'charmm-toppar')
equilibrateSystem(os.path.join(outdir, 'structure.pdb'), os.path.join(outdir, 'structure.psf'), outpdb, charmmfolder=charmmf)
res = Molecule(outpdb)
res.center()
move(os.path.join(outdir, 'structure.pdb'), os.path.join(outdir, 'starting_structure.pdb'))
copy(outpdb, os.path.join(outdir, 'structure.pdb'))
return res
md.run()
# Cleaning up
inputodel = glob(path.join(home(), 'data', 'adaptive', 'input', 'e2*'))
for i in inputodel:
shutil.rmtree(i, ignore_errors=True, acemd='/shared/acemd/bin/acemd')
os.remove(path.join(home(), 'data', 'adaptive', 'input', 'e2_writeinputs.log'))'''
import htmd
import os
import shutil
from htmd.queues.localqueue import LocalGPUQueue
from htmd.simlist import Frame, simlist
from htmd.util import tempname
filedir = htmd.home.home()+'/data/adaptive/'
sims = simlist(glob(os.path.join(filedir, 'data', '*', '')),
glob(os.path.join(filedir, 'input', '*', '')),
glob(os.path.join(filedir, 'input', '*', '')))
outf = tempname()
os.makedirs(outf)
f = Frame(sims[0], 0, 5)
_writeInputsFunction(1, f, 2, outf, 'input.coor')
mol = Molecule(sims[0])
mol.read(os.path.join(outf, 'e2s2_e1s1p0f5', 'input.coor'))
shutil.rmtree(outf)
def loopModeller(mol, segid, seq, startresid, movstart=None, movend=None, modellerexe='mod9.18'):
""" Uses the Modeller software to predict missing loops in a Molecule.
Parameters
----------
mol : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
A Molecule object.
segid : str
The name of the segment containing the gap.
seq : str
The sequence of residues to be added by the loop modeller.
startresid : int
The resid of the residue before the gap.
movstart : int
The resid after which the residues will be allowed to move in modeller.
movend : int
The resid before which the residues will be allowed to move in modeller.
modellerexe : str
The path to the modeller executable.
Returns
-------
newmol : :class:`Molecule <htmd.molecule.molecule.Molecule>` object
A new Molecule object containing the protein with the modelled loop.
Examples
--------
>>> mol = Molecule('1qg8')
>>> mol2 = loopModeller(mol, '0', 'ENR', 133)
"""
if shutil.which(modellerexe) is None:
raise NameError('Could not find a Modeller executable called `{}` in the PATH. This might indicate a wrong path'
' to the executable or a missing installation. To install modeller use `conda install -c '
'salilab modeller` and follow the instructions. To provide the correct path change the '
'`modellerexe` argument'.format(modellerexe))
if movstart is None:
movstart = startresid
if movend is None:
movend = movstart + len(seq) + 1
segatm = mol.segid == segid
segres = np.unique(mol.resid[segatm])
chain = np.unique(mol.chain[segatm])
if len(chain) != 1:
raise RuntimeError('More than one chain detected in segment {}'.format(segid))
pos = np.where(segres == startresid)[0][0]
if (segres[pos+1] - segres[pos]) != (len(seq) + 1):
raise RuntimeError('Sequence is {} characters long while sequence gap ({}-{}) is {} long. Cannot insert loop.'.format(
len(seq), segres[pos], segres[pos+1], (segres[pos+1] - segres[pos])))
chain = chain[0]
segresidstart = np.min(segres)
segresidend = np.max(segres)
currseq = mol.sequence()[segid]
minuses = ''.join(['-'] * len(seq))
inspos = np.where(segres == startresid)[0][0] + 1
gapseq = currseq[:inspos] + minuses + currseq[inspos:] + '*'
fullseq = currseq[:inspos] + seq + currseq[inspos:] + '*'
# Get the sequence of the 1qg8 PDB file, and write to an alignment file
pdbfile = tempname(suffix='.pdb')
mol.write(pdbfile)
alifile = tempname(suffix='.ali')
f = open(alifile, 'w')
f.write('>P1;prot\nstructure:{}:{}:{}:{}:{}::::\n'.format(pdbfile, segresidstart, chain, segresidend, chain))
f.write(gapseq)
f.write('\n>P1;prot_fill\nsequence:::::::::\n')
f.write(fullseq)
f.close()
script = '''
from modeller import *
from modeller.automodel import * # Load the automodel class
log.verbose()
env = environ()
class MyModel(automodel):
def select_atoms(self):
return selection(self.residue_range('{}', '{}'))
a = MyModel(env, alnfile = '{}', knowns = 'prot', sequence = 'prot_fill')
a.starting_model= 1
a.ending_model = 1
a.make()
'''.format(movstart+1-segresidstart, movend-segresidstart, alifile)
pyfile = tempname(suffix='.py')
f = open(pyfile, 'w')
f.write(script)
f.close()
print(pdbfile)
print(alifile)
print(pyfile)
call('{} {}'.format(modellerexe, pyfile), shell=True)
newmol = Molecule('./prot_fill.B99990001.pdb')
print('You can ignore the `import site` error (https://salilab.org/modeller/release.html#issues).'
' The results should have been returned.')
return newmol
