def setupHydrationCalculation(molecule):
"""
"""
# get current directory
current_path = os.getcwd()
# Center the molecule.
OECenter(molecule)
# get molecule name
molecule_name = molecule.GetTitle()
print molecule_name
# create molecule path/directory
molecule_path = os.path.abspath(os.path.join('molecules', molecule_name))
os.makedirs(molecule_path)
# Write mol2 file for the molecule.
writeMolecule(molecule, os.path.join(molecule_path, 'solute.mol2'))
# Write GAFF parameters for gromacs, using antechamber to generate AM1-BCC charges.
parameterizeForGromacs(molecule, topology_filename = os.path.join(molecule_path,'solute.top'), coordinate_filename = os.path.join(molecule_path,'solute.gro'), charge_model = 'bcc', resname = 'MOL')
# Modify gromacs topology file for alchemical free energy calculation.
perturbGromacsTopology(os.path.join(molecule_path,'solute.top'), molecule)
# Convert gromacs topology to an .itp file suitable for inclusion.
top_to_itp(os.path.join(molecule_path,'solute.top'), os.path.join(molecule_path,'solute.itp'), moleculetype = molecule_name)
# SET UP VACUUM SIMULATION
# construct pathname for vacuum simulations
vacuum_path = os.path.join(molecule_path, 'vacuum')
os.mkdir(vacuum_path)
os.chdir(vacuum_path)
# create molecule in enlarged box
command = 'editconf_d -f ../solute.gro -o system.gro -bt cubic -d 50.0 -center 0.0 0.0 0.0'
print command
output = commands.getoutput(command)
print output
# create topology file
topology_contents = """
; amber forcefield
#include "ffamber03.itp"
; my molecule
#include "../solute.itp"
[ system ]
%(molecule_name)s in vacuum
[ molecules ]
; Compound nmols
%(molecule_name)s 1
""" % vars()
writeFile('system.top', topology_contents)
# construct .mdp files
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-vacuum', 'output']))
mdpfile.write(os.path.join(vacuum_path, 'minimization-unconstrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-vacuum', 'constraints', 'output']))
mdpfile.write(os.path.join(vacuum_path, 'minimization-constrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-vacuum', 'constraints', 'thermostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(vacuum_path, 'production.mdp'))
# write shell script for minimization and production
contents = """\
#/bin/tcsh
source /Users/jchodera/local/gromacs-3.1.4-fe/i386-apple-darwin8.11.1/bin/GMXRC.csh
setenv GROMPP grompp_d
setenv MDRUN mdrun_d
# unconstrained minimization
$GROMPP -f minimization-unconstrained.mdp -c system.gro -p system.top -o minimization-unconstrained.tpr -maxwarn 10000
$MDRUN -s minimization-unconstrained.tpr -x minimization-unconstrained.xtc -c minimized-unconstrained.gro -e minimization-unconstrained.edr -g minimization-unconstrained.log
# constrained minimization
$GROMPP -f minimization-constrained.mdp -c minimized-unconstrained.gro -p system.top -o minimization-constrained.tpr -maxwarn 10000
$MDRUN -s minimization-constrained.tpr -x minimization-constrained.xtc -c minimized-constrained.gro -e minimization-constrained.edr -g minimization-constrained.log
# production
$GROMPP -f production.mdp -c minimized-constrained.gro -p system.top -o production.tpr -maxwarn 10000
$MDRUN -v -s production.tpr -o production.trr -x production.xtc -c production.gro -e production.edr -g production.log
echo 0 | trjconv_d -f production.xtc -o production.pdb -s production.tpr
""" % vars()
writeFile(os.path.join(vacuum_path, 'run.sh'), contents)
# SET UP SOLVATED CALCULATION
# construct pathname for solvent simulations
solvated_path = os.path.join(molecule_path, 'solvent')
os.mkdir(solvated_path)
os.chdir(solvated_path)
# write enlarged box for solvent
command = 'editconf_d -f ../solute.gro -o solute.gro -bt octahedron -d 1.0 -center 0.0 0.0 0.0'
print command
output = commands.getoutput(command)
print output
# write topology
topology_contents = """
; amber forcefield
#include "ffamber03.itp"
; my molecule
#include "../solute.itp"
; TIP3P water
#include "ffamber_tip3p.itp"
[ system ]
%(molecule_name)s in solvent
[ molecules ]
; Compound nmols
%(molecule_name)s 1
""" % vars()
writeFile('system.top', topology_contents)
# solvate
command = 'genbox_d -cp solute.gro -cs ffamber_tip3p.gro -o system.gro -p system.top'
# command = 'genbox_d -cp solute.gro -cs tip3p-pme-waterbox.gro -o system.gro -p system.top' # use our special tip3p box
# command = 'genbox_d -cp solute.gro -cs tip3p-pme-waterbox.gro -o system.gro -p system.top -vdwd 0.0' # insert waters but don't cull overlap
print command
output = commands.getoutput(command)
print output
# construct .mdp files
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'output']))
mdpfile.write(os.path.join(solvated_path, 'minimization-unconstrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'constraints', 'output']))
mdpfile.write(os.path.join(solvated_path, 'minimization-constrained.mdp'))
# mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat-equilibration', 'barostat', 'output']))
# mdpfile.randomizeSeed() # randomize velocities
# mdpfile.setParameter('nsteps', '25000') # 10 ps equilibration
# mdpfile.write(os.path.join(solvated_path, 'equilibration.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvated_path, 'production.mdp'))
# write run script
contents = """\
#/bin/tcsh
source /Users/jchodera/local/gromacs-3.1.4-fe/i386-apple-darwin8.11.1/bin/GMXRC.csh
setenv GROMPP grompp_d
setenv MDRUN mdrun_d
# unconstrained minimization
$GROMPP -f minimization-unconstrained.mdp -c system.gro -p system.top -o minimization-unconstrained.tpr -maxwarn 10000
$MDRUN -s minimization-unconstrained.tpr -x minimization-unconstrained.xtc -c minimized-unconstrained.gro -e minimization-unconstrained.edr -g minimization-unconstrained.log
# constrained minimization
$GROMPP -f minimization-constrained.mdp -c minimized-unconstrained.gro -p system.top -o minimization-constrained.tpr -maxwarn 10000
$MDRUN -s minimization-constrained.tpr -x minimization-constrained.xtc -c minimized-constrained.gro -e minimization-constrained.edr -g minimization-constrained.log
# equilibration
#$GROMPP -f equilibration.mdp -c minimized-constrained.gro -p system.top -o equilibration.tpr -maxwarn 10000
#$MDRUN -v -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.gro -e equilibration.edr -g equilibration.log
#echo 0 | trjconv_d -f equilibration.xtc -o equilibration.pdb -s equilibration.tpr
# production
$GROMPP -f production.mdp -c minimized-constrained.gro -p system.top -o production.tpr -maxwarn 10000
#$GROMPP -f production.mdp -c equilibration.gro -p system.top -o production.tpr -maxwarn 10000
#$GROMPP -f production.mdp -c system.gro -p system.top -o production.tpr -maxwarn 10000
$MDRUN -v -s production.tpr -o production.trr -x production.xtc -c production.gro -e production.edr -g production.log
echo 0 | trjconv_d -f production.xtc -o production.pdb -s production.tpr
""" % vars()
writeFile(os.path.join(solvated_path, 'run.sh'), contents)
# restore working directory
os.chdir(current_path)
return
def setupHydrationCalculation(solute,
nreplicates=1,
verbose=True,
jobname="hydration"):
"""Set up an absolute alchemical hydration free energy calculation for the given molecule.
ARGUMENTS
solute (OEMol) - the molecule for which hydration free energy is to be computed (with fully explicit hydrogens) in the desired protonation state.
OPTIONAL ARGUMENTS
nreplicates (integer) - the number of replicates to set up (default: 1)
verbose (boolean) - if True, extra debug information will be printed (default: True)
jobname (string) - string to use for job name (default: "hydration")
NOTES
A directory will be created 'molecules/[molecule name]' as obtained from molecule.GetTitle().
"""
# get current directory
current_path = os.getcwd()
# Center the solute molecule.
OECenter(solute)
# get molecule name
solute_name = molecule.GetTitle()
if verbose: print solute_name
# create molecule path/directory
work_path = os.path.abspath(os.path.join('molecules', solute_name))
os.makedirs(work_path)
# # Write mol2 file for the molecule.
# writeMolecule(solute, os.path.join(solute_path, 'solute.mol2'))
# # Write GAFF parameters for gromacs, using antechamber to generate AM1-BCC charges.
# parameterizeForGromacs(molecule, topology_filename = os.path.join(molecule_path,'solute.top'), coordinate_filename = os.path.join(molecule_path,'solute.gro'), charge_model = 'bcc', resname = 'MOL')
# # Modify gromacs topology file for alchemical free energy calculation.
# perturbGromacsTopology(os.path.join(molecule_path,'solute.top'), molecule)
# # Convert gromacs topology to an .itp file suitable for inclusion.
# top_to_itp(os.path.join(molecule_path,'solute.top'), os.path.join(molecule_path,'solute.itp'), moleculetype = molecule_name)
# get pathnames
mdrun = globals()['mdrun']
grompp = globals()['grompp']
editconf = globals()['editconf']
# SET UP SOLUTE TOPOLOGY
if verbose: print "\nCONSTRUCTING SOLUTE TOPOLOGY"
# Get formal charge of ligand.
solute_charge = formalCharge(solute)
if verbose: print "solute formal charge is %d" % solute_charge
# Write molecule with explicit hydrogens to mol2 file.
print "Writing solute mol2 file..."
solute_mol2_filename = os.path.abspath(
os.path.join(work_path, 'solute.mol2'))
writeMolecule(solute, solute_mol2_filename)
# Set substructure name (which will become residue name).
print "Modifying molecule name..."
modifySubstructureName(solute_mol2_filename, 'MOL')
# Run antechamber to assign GAFF atom types.
print "Running antechamber..."
os.chdir(work_path)
gaff_mol2_filename = os.path.join(work_path, 'solute.gaff.mol2')
charge_model = 'bcc'
command = 'antechamber -i %(solute_mol2_filename)s -fi mol2 -o solute.gaff.mol2 -fo mol2 -c %(charge_model)s -nc %(solute_charge)d > antechamber.out' % vars(
)
if verbose: print command
output = commands.getoutput(command)
if verbose: print output
os.chdir(current_path)
# Generate frcmod file for additional GAFF parameters.
solute_frcmod_filename = os.path.join(work_path, 'frcmod.solute')
command = 'parmchk -i %(gaff_mol2_filename)s -f mol2 -o %(solute_frcmod_filename)s' % vars(
)
if verbose: print command
output = commands.getoutput(command)
if verbose: print output
# Run LEaP to generate topology / coordinates.
solute_prmtop_filename = os.path.join(work_path, 'solute.prmtop')
solute_crd_filename = os.path.join(work_path, 'solute.crd')
solute_off_filename = os.path.join(work_path, 'solute.off')
tleap_input_filename = os.path.join(work_path, 'setup-solute.leap.in')
tleap_output_filename = os.path.join(work_path, 'setup-solute.leap.out')
contents = """
# Load GAFF parameters.
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(solute_frcmod_filename)s
# load solute
solute = loadMol2 %(gaff_mol2_filename)s
# check the solute
check solute
# report net charge
charge solute
# save AMBER parameters
saveAmberParm solute %(solute_prmtop_filename)s %(solute_crd_filename)s
# write .off file
saveOff solute %(solute_off_filename)s
# exit
quit
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# extract total charge
solute_charge = commands.getoutput(
'grep "Total unperturbed charge" %(tleap_output_filename)s | cut -c 27-'
% vars())
solute_charge = int(round(
float(solute_charge))) # round to nearest whole charge
if verbose: print "solute charge is %d" % solute_charge
# PREPARE SOLVATED SOLUTE
print "\nPREPARING SOLVATED SOLUTE"
# create the directory if it doesn't exist
solvent_path = os.path.join(work_path, 'solvent')
if not os.path.exists(solvent_path):
os.makedirs(solvent_path)
# solvate the solute
print "Solvating the solute with tleap..."
system_prmtop_filename = os.path.join(solvent_path, 'system.prmtop')
system_crd_filename = os.path.join(solvent_path, 'system.crd')
tleap_input_filename = os.path.join(solvent_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(solvent_path, 'setup-system.leap.out')
clearance = globals()['clearance'] # clearance around solute (in A)
contents = """
source leaprc.ff99
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(solute_frcmod_filename)s
# Load solute.
loadOff %(solute_off_filename)s
# Create system.
system = combine { solute }
""" % vars()
# add counterions
if (solute_charge != 0):
nions = abs(solute_charge)
if solute_charge < 0: iontype = 'Na+'
if solute_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Solvate in truncated octahedral box.
solvateBox system TIP3PBOX %(clearance)f iso
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters',
'amb2gmx.pl')
system_prefix = os.path.join(solvent_path, 'system')
os.chdir(solvent_path)
command = '%(amb2gmx)s --prmtop system.prmtop --crd system.crd --outname system' % vars(
)
print command
output = commands.getoutput(command)
print output
os.chdir(current_path)
# Extract box size.
g96_filename = os.path.join(solvent_path, 'system.g96')
g96_file = open(g96_filename, 'r')
g96_lines = g96_file.readlines()
g96_file.close()
box_size = zeros([3], float32)
for line_number in range(len(g96_lines)):
if g96_lines[line_number][0:3] == 'BOX':
# parse line with box size in nm
line = g96_lines[line_number + 1]
elements = line.split()
box_size[0] = float(elements[0]) * 10.0
box_size[1] = float(elements[1]) * 10.0
box_size[2] = float(elements[2]) * 10.0
print "box_size = "
print box_size
# make a PDB file for checking
print "Converting system to PDB..."
os.chdir(solvent_path)
command = 'cat system.crd | ambpdb -p system.prmtop > system.pdb' % vars()
output = commands.getoutput(command)
print output
os.chdir(current_path)
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(solvent_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the solute to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
perturbGromacsTopology(system_top_filename,
solute,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices)
# set up replicates
for replicate in range(nreplicates):
# Create replicate directory.
working_path = os.path.join(solvent_path, '%d' % replicate)
os.makedirs(working_path)
# TODO: Modify solute coordinates in system.g96 to cycle through available conformations.
# set up mdp files
print "Writing mdp files for replicate %d..." % replicate
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-solvent', 'constraints',
'output'
]))
mdpfile.write(os.path.join(working_path, 'minimize.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'thermostat', 'barostat', 'output'
]))
mdpfile.setParameter('nsteps', '10000') # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(working_path, 'equilibration.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'thermostat', 'barostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(working_path, 'production.mdp'))
# write run script
print "Writing run script..."
solvent_path = os.path.abspath(solvent_path)
contents = """\
set#!/bin/tcsh
#BSUB -J %(jobname)s_solvent
#BSUB -o %(working_path)s/outfile.%%J
#BSUB -e %(working_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source $GMXRC
cd %(working_path)s
# constrained minimize
%(grompp)s -f minimize.mdp -c ../system.g96 -p ../system.top -o minimize.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# equilibration
%(grompp)s -f equilibration.mdp -c minimize.g96 -p ../system.top -o equilibration.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# production
%(grompp)s -f production.mdp -c equilibration.g96 -p ../system.top -o production.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(working_path, 'run.sh'), contents)
# SET UP VACUUM SIMULATION
# construct pathname for vacuum simulations
vacuum_path = os.path.join(work_path, 'vacuum')
if not os.path.exists(vacuum_path):
os.makedirs(vacuum_path)
# solvate the solute
print "Preparing vacuum solute with tleap..."
system_prmtop_filename = os.path.join(vacuum_path, 'system.prmtop')
system_crd_filename = os.path.join(vacuum_path, 'system.crd')
tleap_input_filename = os.path.join(vacuum_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(vacuum_path, 'setup-system.leap.out')
clearance = 50.0 # clearance in A
contents = """
source leaprc.ff99
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(solute_frcmod_filename)s
# Load solute.
loadOff %(solute_off_filename)s
# Create system.
system = combine { solute }
""" % vars()
# add counterions
if (solute_charge != 0):
nions = abs(solute_charge)
if solute_charge < 0: iontype = 'Na+'
if solute_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Create big box.
setBox system centers %(clearance)f iso
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters',
'amb2gmx.pl')
os.chdir(vacuum_path)
command = '%(amb2gmx)s --prmtop system.prmtop --crd system.crd --outname system' % vars(
)
print command
output = commands.getoutput(command)
print output
os.chdir(current_path)
# make a PDB file for checking
print "Converting system to PDB..."
os.chdir(vacuum_path)
command = 'cat system.crd | ambpdb -p system.prmtop > system.pdb' % vars()
output = commands.getoutput(command)
print output
os.chdir(current_path)
# write enlarged box for solvent because LEaP doesn't do it right.
system_g96_filename = os.path.join(vacuum_path, 'system.g96')
command = '%s -f %s -o %s -bt cubic -box %f %f %f -center 0.0 0.0 0.0' % (
editconf, system_g96_filename, system_g96_filename, box_size[0] / 10.0,
box_size[1] / 10.0, box_size[2] / 10.0)
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(vacuum_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
nions = 0
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# DEBUG
# add those atoms in the solute to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
# add counterions to balance solute
if ((atom['residue'] == 'Na+') or
(atom['residue'] == 'Cl-')) and (nions < abs(solute_charge)):
print "WARNING: Solute has net charge of %(solute_charge)d -- will perturb counterions too." % vars(
)
perturb_atom_indices.append(atom['nr'])
nions += 1
perturbGromacsTopology(system_top_filename,
solute,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices)
# construct replicates
for replicate in range(nreplicates):
# Create replicate path.
working_path = os.path.join(vacuum_path, '%d' % replicate)
os.makedirs(working_path)
# TODO: Modify solute coordinates in system.g96 to cycle through available conformations.
# construct .mdp files
print "Writing mdp files for replicate %d..." % replicate
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-vacuum', 'constraints',
'output'
]))
cutoff = box_size.min() / 10.0 / 2.0 - 0.0001 # compute cutoff
mdpfile.setParameter('rlist', '%f' % cutoff)
mdpfile.setParameter('rvdw', '%f' % cutoff)
mdpfile.setParameter('rcoulomb', '%f' % cutoff)
mdpfile.write(os.path.join(working_path, 'minimize.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-vacuum', 'constraints',
'thermostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.setParameter('rlist', '%f' % cutoff)
mdpfile.setParameter('rvdw', '%f' % cutoff)
mdpfile.setParameter('rcoulomb', '%f' % cutoff)
mdpfile.write(os.path.join(working_path, 'production.mdp'))
# write shell script for minimization and production
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_vacuum
#BSUB -o %(working_path)s/outfile.%%J
#BSUB -e %(working_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source $GMXRC
cd %(working_path)s
# constrained minimize
%(grompp)s -f minimize.mdp -c ../system.g96 -p ../system.top -o minimize.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# production
%(grompp)s -f production.mdp -c minimize.g96 -p ../system.top -o production.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(working_path, 'run.sh'), contents)
return
def setup_system(protein_pdb_filename, ligand_filename, work_path, parameter_path, GMXRC, jobname):
"""Set up a system for alchemical free energy calculation in gromacs.
ARGUMENTS
protein_pdb_filename (string) - name of ffamber-named protein PDB file
ligand_filename (string) - name of mol2 file describing ligand
work_path (string) - name of directory to place files in
"""
# get current directory
current_path = os.getcwd()
# create the work directory if it doesn't exist
if not os.path.exists(work_path):
os.makedirs(work_path)
mdrun = globals()['mdrun']
grompp = globals()['grompp']
# SET UP PROTEIN TOPOLOGY
print "\nCONSTRUCTING PROTEIN TOPOLOGY"
# convert protein PDB file to AMBER naming conventions, dropping atoms that AMBER won't recognize (like protons)
print "Converting PDB naming to AMBER..."
protein_amberpdb_filename = os.path.join(work_path, 'protein-ambernames.pdb')
rename_pdb_for_amber(protein_pdb_filename, protein_amberpdb_filename)
# run leap to set up protein and report on net charge
print "Running LEaP to set up protein..."
protein_prmtop_filename = os.path.join(work_path,'protein.prmtop')
protein_crd_filename = os.path.join(work_path,'protein.crd')
protein_off_filename = os.path.join(work_path, 'protein.off')
tleap_input_filename = os.path.join(work_path, 'setup-protein.leap.in')
tleap_output_filename = os.path.join(work_path, 'setup-protein.leap.out')
contents = """
# Load AMBER ff99 parameters.
source leaprc.ff99
# Load phosphoresidue parameters.
loadoff %(parameter_path)s/T1P.off
loadoff %(parameter_path)s/T2P.off
loadoff %(parameter_path)s/Y1P.off
loadoff %(parameter_path)s/Y2P.off
loadamberparams %(parameter_path)s/frcmod_t1p
loadamberparams %(parameter_path)s/frcmod_t2p
loadamberparams %(parameter_path)s/frcmod_y1p
loadamberparams %(parameter_path)s/frcmod_y2p
# Load PDB file with AMBER naming conventions, stripped of hydrogens.
protein = loadpdb %(protein_amberpdb_filename)s
# check the system
check protein
# report net charge
charge protein
# write out parameters
saveAmberParm protein %(protein_prmtop_filename)s %(protein_crd_filename)s
# write as LEaP object
saveOff protein %(protein_off_filename)s
# exit
quit
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars()
output = commands.getoutput(command)
# extract total charge
protein_charge = commands.getoutput('grep "Total unperturbed charge" %(tleap_output_filename)s | cut -c 27-' % vars())
protein_charge = int(round(float(protein_charge))) # round to nearest whole charge
print protein_charge
# SET UP LIGAND TOPOLOGY
print "\nCONSTRUCTING LIGAND TOPOLOGY"
# Read the ligand into OEMol.
print "Reading ligand..."
ligand = readMolecule(ligand_filename, normalize = True)
# Get formal charge of ligand.
ligand_charge = formalCharge(ligand)
print "formal charge is %d" % ligand_charge
# TODO: Choose most appropriate protonation and tautomeric state.
# Write molecule with explicit hydrogens to mol2 file.
print "Writing ligand mol2 file..."
ligand_mol2_filename = os.path.abspath(os.path.join(work_path, 'ligand.mol2'))
writeMolecule(ligand, ligand_mol2_filename)
# Set substructure name (which will become residue name).
print "Modifying molecule name..."
modifySubstructureName(ligand_mol2_filename, 'MOL')
# Run antechamber to assign GAFF atom types.
print "Running antechamber..."
os.chdir(work_path)
gaff_mol2_filename = os.path.join(work_path, 'ligand.gaff.mol2')
charge_model = 'bcc'
command = 'antechamber -i %(ligand_mol2_filename)s -fi mol2 -o ligand.gaff.mol2 -fo mol2 -c %(charge_model)s -nc %(ligand_charge)d > antechamber.out' % vars()
print command
output = commands.getoutput(command)
os.chdir(current_path)
# Generate frcmod file for additional GAFF parameters.
ligand_frcmod_filename = os.path.join(work_path, 'frcmod.ligand')
output = commands.getoutput('parmchk -i %(gaff_mol2_filename)s -f mol2 -o %(ligand_frcmod_filename)s' % vars())
print output
# Run LEaP to generate topology / coordinates.
ligand_prmtop_filename = os.path.join(work_path,'ligand.prmtop')
ligand_crd_filename = os.path.join(work_path,'ligand.crd')
ligand_off_filename = os.path.join(work_path, 'ligand.off')
tleap_input_filename = os.path.join(work_path, 'setup-ligand.leap.in')
tleap_output_filename = os.path.join(work_path, 'setup-ligand.leap.out')
contents = """
# Load GAFF parameters.
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# load ligand
ligand = loadMol2 %(gaff_mol2_filename)s
# check the ligand
check ligand
# report net charge
charge ligand
# save AMBER parameters
saveAmberParm ligand %(ligand_prmtop_filename)s %(ligand_crd_filename)s
# write .off file
saveOff ligand %(ligand_off_filename)s
# exit
quit
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars()
output = commands.getoutput(command)
# extract total charge
ligand_charge = commands.getoutput('grep "Total unperturbed charge" %(tleap_output_filename)s | cut -c 27-' % vars())
ligand_charge = int(round(float(ligand_charge))) # round to nearest whole charge
print "ligand charge is %d" % ligand_charge
# SET UP VACUUM SIMULATION
# construct pathname for vacuum simulations
vacuum_path = os.path.join(work_path, 'vacuum')
if not os.path.exists(vacuum_path):
os.makedirs(vacuum_path)
# solvate the ligand
print "Preparing vacuum ligand with tleap..."
system_prmtop_filename = os.path.join(vacuum_path,'system.prmtop')
system_crd_filename = os.path.join(vacuum_path,'system.crd')
tleap_input_filename = os.path.join(vacuum_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(vacuum_path, 'setup-system.leap.out')
clearance = 50.0 # clearance in A
contents = """
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { ligand }
""" % vars()
# add counterions
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Create big box.
setBox system centers %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars()
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters', 'amb2gmx.pl')
system_prefix = os.path.join(vacuum_path, 'system')
command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars()
print command
output = commands.getoutput(command)
print output
# write enlarged box for solvent because LEaP doesn't do it right.
system_g96_filename = os.path.join(vacuum_path, 'system.g96')
command = '%s -f %s -o %s -bt octahedron -d %f -center 0.0 0.0 0.0' % (editconf, system_g96_filename, system_g96_filename, clearance)
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(vacuum_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
nions = 0
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the ligand to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
# add counterions to balance ligand
if ((atom['residue'] == 'Na+') or (atom['residue'] == 'Cl-')) and (nions < abs(ligand_charge)):
perturb_atom_indices.append(atom['nr'])
nions += 1
perturbGromacsTopology(system_top_filename, ligand, perturb_torsions = True, perturb_vdw = True, perturb_charges = True, perturb_atom_indices = perturb_atom_indices)
# construct .mdp files
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-vacuum', 'constraints', 'output']))
mdpfile.write(os.path.join(vacuum_path, 'minimize.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-vacuum', 'constraints', 'thermostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(vacuum_path, 'production.mdp'))
vacuum_path = os.path.abspath(vacuum_path)
# write shell script for minimization and production
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_vacuum
#BSUB -o %(vacuum_path)s/outfile.%%J
#BSUB -e %(vacuum_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source %(GMXRC)s
cd %(vacuum_path)s
setenv RANSEED 1234
# constrained minimize
%(grompp)s -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# production
%(grompp)s -f production.mdp -c minimize.g96 -p system.top -o production.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(vacuum_path, 'run.sh'), contents)
# PREPARE SOLVATED LIGAND
print "\nPREPARING SOLVATED LIGAND"
# create the directory if it doesn't exist
solvent_path = os.path.join(work_path, 'solvent')
if not os.path.exists(solvent_path):
os.makedirs(solvent_path)
# solvate the ligand
print "Solvating the ligand with tleap..."
system_prmtop_filename = os.path.join(solvent_path,'system.prmtop')
system_crd_filename = os.path.join(solvent_path,'system.crd')
tleap_input_filename = os.path.join(solvent_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(solvent_path, 'setup-system.leap.out')
clearance = 10.0 # clearance in A
contents = """
source leaprc.ff99
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { ligand }
""" % vars()
# add counterions
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Solvate in truncated octahedral box.
solvateOct system TIP3PBOX %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars()
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters', 'amb2gmx.pl')
system_prefix = os.path.join(solvent_path, 'system')
command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars()
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(solvent_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the ligand to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
perturbGromacsTopology(system_top_filename, ligand, perturb_torsions = True, perturb_vdw = True, perturb_charges = True, perturb_atom_indices = perturb_atom_indices)
# set up mdp files
print "Writing mdp files..."
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'constraints', 'output']))
mdpfile.write(os.path.join(solvent_path, 'minimize.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'output']))
mdpfile.setParameter('nsteps', '10000') # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvent_path, 'equilibration.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvent_path, 'production.mdp'))
# write run script
print "Writing run script..."
solvent_path = os.path.abspath(solvent_path)
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_solvent
#BSUB -o %(solvent_path)s/outfile.%%J
#BSUB -e %(solvent_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source %(GMXRC)s
cd %(solvent_path)s
setenv RANSEED 1234
# constrained minimize
%(grompp)s -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# equilibration
%(grompp)s -f equilibration.mdp -c minimize.g96 -p system.top -o equilibration.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# production
%(grompp)s -f production.mdp -c equilibration.g96 -p system.top -o production.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(solvent_path, 'run.sh'), contents)
# PREPARE SOLVATED COMPLEX
print "\nPREPARING SOLVATED COMPLEX"
# create the directory if it doesn't exist
complex_path = os.path.join(work_path, 'complex')
if not os.path.exists(complex_path):
os.makedirs(complex_path)
# solvate the ligand
print "Solvating the complex with tleap..."
system_prmtop_filename = os.path.join(complex_path,'system.prmtop')
system_crd_filename = os.path.join(complex_path,'system.crd')
tleap_input_filename = os.path.join(complex_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(complex_path, 'setup-system.leap.out')
clearance = 10.0 # clearance in A
contents = """
source leaprc.ff99
source leaprc.gaff
loadamberparams %(parameter_path)s/frcmod_t1p
loadamberparams %(parameter_path)s/frcmod_t2p
loadamberparams %(parameter_path)s/frcmod_y1p
loadamberparams %(parameter_path)s/frcmod_y2p
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load protein.
loadOff %(protein_off_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { protein ligand }
""" % vars()
# add counterions for ligand
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions for ligand (will be annihilated with ligand).
addions system %(iontype)s %(nions)d
""" % vars()
#
# add counterions for protein
if (protein_charge != 0):
nions = abs(protein_charge)
if protein_charge < 0: iontype = 'Na+'
if protein_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions for protein.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Solvate in truncated octahedral box.
solvateOct system TIP3PBOX %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars()
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters', 'amb2gmx.pl')
system_prefix = os.path.join(complex_path, 'system')
command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars()
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(complex_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
nions = 0
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the ligand to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
# add counterions to balance ligand
if ((atom['residue'] == 'Na+') or (atom['residue'] == 'Cl-')) and (nions < abs(ligand_charge)):
perturb_atom_indices.append(atom['nr'])
nions += 1
perturbGromacsTopology(system_top_filename, ligand, perturb_torsions = True, perturb_vdw = True, perturb_charges = True, perturb_atom_indices = perturb_atom_indices)
# set up mdp files
print "Writing mdp files..."
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'constraints', 'restraints', 'output']))
mdpfile.write(os.path.join(complex_path, 'minimize.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'output']))
mdpfile.setParameter('nsteps', '10000') # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'equilibration.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'restraints', 'thermostat', 'barostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'production.mdp'))
# Copy restraint scripts to run directory
shutil.copy(compute_angles,complex_path)
shutil.copy(restraint_topology,complex_path)
# write run script
print "Writing run script..."
complex_path = os.path.abspath(complex_path)
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_complex
#BSUB -o %(complex_path)s/outfile.%%J
#BSUB -e %(complex_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source %(GMXRC)s
cd %(complex_path)s
setenv RANSEED 1234
# create a tpr file from which to create restraints
%(grompp)s -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
# Integrate restraints into pre-minimization topfile
python computeangles.py -f system.g96 -s minimize.tpr -d $RANSEED
python restraint_topology.py -n system -p .
# constrained minimize with restraints
%(grompp)s -f minimize.mdp -c system_restr.g96 -p system_restr.top -o minimize.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# equilibration with restraints
%(grompp)s -f equilibration.mdp -c minimize.g96 -p system.top -o equilibration.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# create a TPR file, integrate restraints
%(grompp)s -f production.mdp -c equilibration.g96 -p system_restr.top -o production.tpr -maxwarn 10000 -n system.ndx
cp system_restr.top minimize.top
python computeangles.py -f equilibration.g96 -s production.tpr -d $RANSEED
python restraint_topology.py -n equilibration -p .
# production with restraints
%(grompp)s -f production.mdp -c equilibration_restr.g96 -p minimize_restr.top -o production.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(complex_path, 'run.sh'), contents)
return
def setup_system(protein_pdb_filename, ligand_filename, work_path,
parameter_path, GMXRC, jobname):
"""Set up a system for alchemical free energy calculation in gromacs.
ARGUMENTS
protein_pdb_filename (string) - name of ffamber-named protein PDB file
ligand_filename (string) - name of mol2 file describing ligand
work_path (string) - name of directory to place files in
"""
# get current directory
current_path = os.getcwd()
# create the work directory if it doesn't exist
if not os.path.exists(work_path):
os.makedirs(work_path)
mdrun = globals()['mdrun']
grompp = globals()['grompp']
# SET UP PROTEIN TOPOLOGY
print "\nCONSTRUCTING PROTEIN TOPOLOGY"
# convert protein PDB file to AMBER naming conventions, dropping atoms that AMBER won't recognize (like protons)
print "Converting PDB naming to AMBER..."
protein_amberpdb_filename = os.path.join(work_path,
'protein-ambernames.pdb')
rename_pdb_for_amber(protein_pdb_filename, protein_amberpdb_filename)
# run leap to set up protein and report on net charge
print "Running LEaP to set up protein..."
protein_prmtop_filename = os.path.join(work_path, 'protein.prmtop')
protein_crd_filename = os.path.join(work_path, 'protein.crd')
protein_off_filename = os.path.join(work_path, 'protein.off')
tleap_input_filename = os.path.join(work_path, 'setup-protein.leap.in')
tleap_output_filename = os.path.join(work_path, 'setup-protein.leap.out')
contents = """
# Load AMBER ff99 parameters.
source leaprc.ff99
# Load phosphoresidue parameters.
loadoff %(parameter_path)s/T1P.off
loadoff %(parameter_path)s/T2P.off
loadoff %(parameter_path)s/Y1P.off
loadoff %(parameter_path)s/Y2P.off
loadamberparams %(parameter_path)s/frcmod_t1p
loadamberparams %(parameter_path)s/frcmod_t2p
loadamberparams %(parameter_path)s/frcmod_y1p
loadamberparams %(parameter_path)s/frcmod_y2p
# Load PDB file with AMBER naming conventions, stripped of hydrogens.
protein = loadpdb %(protein_amberpdb_filename)s
# check the system
check protein
# report net charge
charge protein
# write out parameters
saveAmberParm protein %(protein_prmtop_filename)s %(protein_crd_filename)s
# write as LEaP object
saveOff protein %(protein_off_filename)s
# exit
quit
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# extract total charge
protein_charge = commands.getoutput(
'grep "Total unperturbed charge" %(tleap_output_filename)s | cut -c 27-'
% vars())
protein_charge = int(round(
float(protein_charge))) # round to nearest whole charge
print protein_charge
# SET UP LIGAND TOPOLOGY
print "\nCONSTRUCTING LIGAND TOPOLOGY"
# Read the ligand into OEMol.
print "Reading ligand..."
ligand = readMolecule(ligand_filename, normalize=True)
# Get formal charge of ligand.
ligand_charge = formalCharge(ligand)
print "formal charge is %d" % ligand_charge
# TODO: Choose most appropriate protonation and tautomeric state.
# Write molecule with explicit hydrogens to mol2 file.
print "Writing ligand mol2 file..."
ligand_mol2_filename = os.path.abspath(
os.path.join(work_path, 'ligand.mol2'))
writeMolecule(ligand, ligand_mol2_filename)
# Set substructure name (which will become residue name).
print "Modifying molecule name..."
modifySubstructureName(ligand_mol2_filename, 'MOL')
# Run antechamber to assign GAFF atom types.
print "Running antechamber..."
os.chdir(work_path)
gaff_mol2_filename = os.path.join(work_path, 'ligand.gaff.mol2')
charge_model = 'bcc'
command = 'antechamber -i %(ligand_mol2_filename)s -fi mol2 -o ligand.gaff.mol2 -fo mol2 -c %(charge_model)s -nc %(ligand_charge)d > antechamber.out' % vars(
)
print command
output = commands.getoutput(command)
os.chdir(current_path)
# Generate frcmod file for additional GAFF parameters.
ligand_frcmod_filename = os.path.join(work_path, 'frcmod.ligand')
output = commands.getoutput(
'parmchk -i %(gaff_mol2_filename)s -f mol2 -o %(ligand_frcmod_filename)s'
% vars())
print output
# Run LEaP to generate topology / coordinates.
ligand_prmtop_filename = os.path.join(work_path, 'ligand.prmtop')
ligand_crd_filename = os.path.join(work_path, 'ligand.crd')
ligand_off_filename = os.path.join(work_path, 'ligand.off')
tleap_input_filename = os.path.join(work_path, 'setup-ligand.leap.in')
tleap_output_filename = os.path.join(work_path, 'setup-ligand.leap.out')
contents = """
# Load GAFF parameters.
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# load ligand
ligand = loadMol2 %(gaff_mol2_filename)s
# check the ligand
check ligand
# report net charge
charge ligand
# save AMBER parameters
saveAmberParm ligand %(ligand_prmtop_filename)s %(ligand_crd_filename)s
# write .off file
saveOff ligand %(ligand_off_filename)s
# exit
quit
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# extract total charge
ligand_charge = commands.getoutput(
'grep "Total unperturbed charge" %(tleap_output_filename)s | cut -c 27-'
% vars())
ligand_charge = int(round(
float(ligand_charge))) # round to nearest whole charge
print "ligand charge is %d" % ligand_charge
# SET UP VACUUM SIMULATION
# construct pathname for vacuum simulations
vacuum_path = os.path.join(work_path, 'vacuum')
if not os.path.exists(vacuum_path):
os.makedirs(vacuum_path)
# solvate the ligand
print "Preparing vacuum ligand with tleap..."
system_prmtop_filename = os.path.join(vacuum_path, 'system.prmtop')
system_crd_filename = os.path.join(vacuum_path, 'system.crd')
tleap_input_filename = os.path.join(vacuum_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(vacuum_path, 'setup-system.leap.out')
clearance = 50.0 # clearance in A
contents = """
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { ligand }
""" % vars()
# add counterions
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Create big box.
setBox system centers %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters',
'amb2gmx.pl')
system_prefix = os.path.join(vacuum_path, 'system')
command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars(
)
print command
output = commands.getoutput(command)
print output
# write enlarged box for solvent because LEaP doesn't do it right.
system_g96_filename = os.path.join(vacuum_path, 'system.g96')
command = '%s -f %s -o %s -bt octahedron -d %f -center 0.0 0.0 0.0' % (
editconf, system_g96_filename, system_g96_filename, clearance)
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(vacuum_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
nions = 0
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the ligand to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
# add counterions to balance ligand
if ((atom['residue'] == 'Na+') or
(atom['residue'] == 'Cl-')) and (nions < abs(ligand_charge)):
perturb_atom_indices.append(atom['nr'])
nions += 1
perturbGromacsTopology(system_top_filename,
ligand,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices)
# construct .mdp files
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-vacuum', 'constraints',
'output'
]))
mdpfile.write(os.path.join(vacuum_path, 'minimize.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-vacuum', 'constraints',
'thermostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(vacuum_path, 'production.mdp'))
vacuum_path = os.path.abspath(vacuum_path)
# write shell script for minimization and production
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_vacuum
#BSUB -o %(vacuum_path)s/outfile.%%J
#BSUB -e %(vacuum_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source %(GMXRC)s
cd %(vacuum_path)s
setenv RANSEED 1234
# constrained minimize
%(grompp)s -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# production
%(grompp)s -f production.mdp -c minimize.g96 -p system.top -o production.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(vacuum_path, 'run.sh'), contents)
# PREPARE SOLVATED LIGAND
print "\nPREPARING SOLVATED LIGAND"
# create the directory if it doesn't exist
solvent_path = os.path.join(work_path, 'solvent')
if not os.path.exists(solvent_path):
os.makedirs(solvent_path)
# solvate the ligand
print "Solvating the ligand with tleap..."
system_prmtop_filename = os.path.join(solvent_path, 'system.prmtop')
system_crd_filename = os.path.join(solvent_path, 'system.crd')
tleap_input_filename = os.path.join(solvent_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(solvent_path, 'setup-system.leap.out')
clearance = 10.0 # clearance in A
contents = """
source leaprc.ff99
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { ligand }
""" % vars()
# add counterions
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Solvate in truncated octahedral box.
solvateOct system TIP3PBOX %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters',
'amb2gmx.pl')
system_prefix = os.path.join(solvent_path, 'system')
command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars(
)
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(solvent_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the ligand to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
perturbGromacsTopology(system_top_filename,
ligand,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices)
# set up mdp files
print "Writing mdp files..."
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-solvent', 'constraints',
'output'
]))
mdpfile.write(os.path.join(solvent_path, 'minimize.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'thermostat', 'barostat', 'output'
]))
mdpfile.setParameter('nsteps', '10000') # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvent_path, 'equilibration.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'thermostat', 'barostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvent_path, 'production.mdp'))
# write run script
print "Writing run script..."
solvent_path = os.path.abspath(solvent_path)
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_solvent
#BSUB -o %(solvent_path)s/outfile.%%J
#BSUB -e %(solvent_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source %(GMXRC)s
cd %(solvent_path)s
setenv RANSEED 1234
# constrained minimize
%(grompp)s -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# equilibration
%(grompp)s -f equilibration.mdp -c minimize.g96 -p system.top -o equilibration.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# production
%(grompp)s -f production.mdp -c equilibration.g96 -p system.top -o production.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(solvent_path, 'run.sh'), contents)
# PREPARE SOLVATED COMPLEX
print "\nPREPARING SOLVATED COMPLEX"
# create the directory if it doesn't exist
complex_path = os.path.join(work_path, 'complex')
if not os.path.exists(complex_path):
os.makedirs(complex_path)
# solvate the ligand
print "Solvating the complex with tleap..."
system_prmtop_filename = os.path.join(complex_path, 'system.prmtop')
system_crd_filename = os.path.join(complex_path, 'system.crd')
tleap_input_filename = os.path.join(complex_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(complex_path, 'setup-system.leap.out')
clearance = 10.0 # clearance in A
contents = """
source leaprc.ff99
source leaprc.gaff
loadamberparams %(parameter_path)s/frcmod_t1p
loadamberparams %(parameter_path)s/frcmod_t2p
loadamberparams %(parameter_path)s/frcmod_y1p
loadamberparams %(parameter_path)s/frcmod_y2p
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load protein.
loadOff %(protein_off_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { protein ligand }
""" % vars()
# add counterions for ligand
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions for ligand (will be annihilated with ligand).
addions system %(iontype)s %(nions)d
""" % vars()
#
# add counterions for protein
if (protein_charge != 0):
nions = abs(protein_charge)
if protein_charge < 0: iontype = 'Na+'
if protein_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions for protein.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Solvate in truncated octahedral box.
solvateOct system TIP3PBOX %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars(
)
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters',
'amb2gmx.pl')
system_prefix = os.path.join(complex_path, 'system')
command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars(
)
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(complex_path, 'system.top')
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, 'atoms')
nions = 0
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if (nelements < 8): continue
# parse line
atom = dict()
atom['nr'] = int(elements[0])
atom['type'] = elements[1]
atom['resnr'] = int(elements[2])
atom['residue'] = elements[3]
atom['atom'] = elements[4]
atom['cgnr'] = int(elements[5])
atom['charge'] = float(elements[6])
atom['mass'] = float(elements[7])
# add those atoms in the ligand to our list
if atom['residue'] == 'MOL':
perturb_atom_indices.append(atom['nr'])
# add counterions to balance ligand
if ((atom['residue'] == 'Na+') or
(atom['residue'] == 'Cl-')) and (nions < abs(ligand_charge)):
perturb_atom_indices.append(atom['nr'])
nions += 1
perturbGromacsTopology(system_top_filename,
ligand,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices)
# set up mdp files
print "Writing mdp files..."
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-solvent', 'constraints',
'restraints', 'output'
]))
mdpfile.write(os.path.join(complex_path, 'minimize.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'thermostat', 'barostat', 'output'
]))
mdpfile.setParameter('nsteps', '10000') # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'equilibration.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'restraints', 'thermostat', 'barostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'production.mdp'))
# Copy restraint scripts to run directory
shutil.copy(compute_angles, complex_path)
shutil.copy(restraint_topology, complex_path)
# write run script
print "Writing run script..."
complex_path = os.path.abspath(complex_path)
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_complex
#BSUB -o %(complex_path)s/outfile.%%J
#BSUB -e %(complex_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source %(GMXRC)s
cd %(complex_path)s
setenv RANSEED 1234
# create a tpr file from which to create restraints
%(grompp)s -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
# Integrate restraints into pre-minimization topfile
python computeangles.py -f system.g96 -s minimize.tpr -d $RANSEED
python restraint_topology.py -n system -p .
# constrained minimize with restraints
%(grompp)s -f minimize.mdp -c system_restr.g96 -p system_restr.top -o minimize.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# equilibration with restraints
%(grompp)s -f equilibration.mdp -c minimize.g96 -p system.top -o equilibration.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# create a TPR file, integrate restraints
%(grompp)s -f production.mdp -c equilibration.g96 -p system_restr.top -o production.tpr -maxwarn 10000 -n system.ndx
cp system_restr.top minimize.top
python computeangles.py -f equilibration.g96 -s production.tpr -d $RANSEED
python restraint_topology.py -n equilibration -p .
# production with restraints
%(grompp)s -f production.mdp -c equilibration_restr.g96 -p minimize_restr.top -o production.tpr -maxwarn 10000 -n system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(complex_path, 'run.sh'), contents)
return
def setup_complex_simulation(complex_path, jobname, ligand, ligand_off_filename, ligand_frcmod_filename, protein_off_filename, protein_charge, common_substructure):
"""Set up free energy calculation for ligand in complex.
ARGUMENTS
complex_path (string) - the pathname where the simulation is to be set up (created if doesn't exist).
jobname (string) - job name to be used to form batch queue job name
ligand (OEMol) - the ligand
ligand_off_filename (string) - leap library for ligand
ligand_frcmod_filename (string) - additional ligand parameters
common_substructure (OEMol) -
"""
print "\nPREPARING SOLVATED COMPLEX"
# create path if it doesn't exist
if not os.path.exists(complex_path):
os.makedirs(complex_path)
# get ligand formal charge
ligand_charge = formalCharge(ligand)
# set up the system in tLEaP
print "Solvating the complex with tleap..."
system_prmtop_filename = os.path.join(complex_path,'system.prmtop')
system_crd_filename = os.path.join(complex_path,'system.crd')
tleap_input_filename = os.path.join(complex_path, 'setup-system.leap.in')
tleap_output_filename = os.path.join(complex_path, 'setup-system.leap.out')
clearance = 10.0 # clearance in A to add around protein
contents = """
source leaprc.ff03
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(ligand_frcmod_filename)s
# Load protein.
loadOff %(protein_off_filename)s
# Load ligand.
loadOff %(ligand_off_filename)s
# Create system.
system = combine { protein ligand }
""" % vars()
# add counterions for ligand
if (ligand_charge != 0):
nions = abs(ligand_charge)
if ligand_charge < 0: iontype = 'Na+'
if ligand_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions for ligand (will be annihilated with ligand).
addions system %(iontype)s %(nions)d
""" % vars()
#
# add counterions for protein
if (protein_charge != 0):
nions = abs(protein_charge)
if protein_charge < 0: iontype = 'Na+'
if protein_charge > 0: iontype = 'Cl-'
contents += """
# Add counterions for protein.
addions system %(iontype)s %(nions)d
""" % vars()
#
contents += """
# Solvate in truncated octahedral box.
# solvateOct system TIP3PBOX %(clearance)f
# solvate in rectilinear box
solvateBox system TIP3PBOX %(clearance)f
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
""" % vars()
write_file(tleap_input_filename, contents)
command = 'tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s' % vars()
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv('MMTOOLSPATH'), 'converters', 'amb2gmx.pl')
system_prefix = os.path.join(complex_path, 'system')
current_path = os.getcwd()
os.chdir(complex_path)
#command = '%(amb2gmx)s --prmtop %(system_prmtop_filename)s --crd %(system_crd_filename)s --outname %(system_prefix)s' % vars()
command = '%(amb2gmx)s --prmtop system.prmtop --crd system.crd --outname system' % vars()
print command
output = commands.getoutput(command)
print output
os.chdir(current_path)
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(complex_path, 'system.top')
perturbGromacsTopologyToIntermediate(system_top_filename, 'MOL', ligand, common_substructure, perturb_torsions = True)
# set up mdp files
print "Writing mdp files..."
mdpfile = MdpFile(compose_blocks(['header', 'minimizationMZ', 'nonbonded-solvent', 'constraints', 'restraints', 'output']))
mdpfile.write(os.path.join(complex_path, 'minimize.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'output']))
mdpfile.setParameter('nsteps', '10000') # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'equilibration.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'restraints', 'thermostat', 'barostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'production.mdp'))
# Copy restraint scripts to run directory
shutil.copy(os.path.join(script_basepath,'compute_angles.py'),complex_path)
shutil.copy(os.path.join(script_basepath,'restraint_topology.py'),complex_path)
# write batch queue script
print "Writing batch queue script..."
complex_path = os.path.abspath(complex_path)
contents = """\
#!/bin/tcsh
#BSUB -J %(jobname)s_complex
#BSUB -n 1
#BSUB -M 2000000
source $GMXRC
# create a tpr file from which to create restraints
grompp -f minimize.mdp -c system.g96 -p system.top -o minimize.tpr -maxwarn 10000 -n system.ndx
# Integrate restraints into pre-minimization topfile
python compute_angles.py -f system.g96 -s minimize.tpr -d $RANSEED
python restraint_topology.py -n system -p .
# constrained minimize with restraints
grompp -f minimize.mdp -c system_restr.g96 -p system_restr.top -o minimize.tpr -maxwarn 10000 -n system.ndx
mdrun -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# create a tpr file from which to create restraints
grompp -f equilibration.mdp -c minimize.g96 -p system_restr.top -o equilibration.tpr -maxwarn 10000 -n system.ndx
# Integrate restraints into pre-equilibration topfile
cp system_restr.top minimize.top
#python compute_angles.py -f minimize.g96 -s equilibration.tpr -d $RANSEED
python restraint_topology.py -n minimize -p .
# equilibration with restraints
grompp -f equilibration.mdp -c minimize_restr.g96 -p minimize_restr.top -o equilibration.tpr -maxwarn 10000 -n system.ndx
mdrun -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# create a TPR file, integrate restraints
grompp -f production.mdp -c equilibration.g96 -p minimize_restr.top -o production.tpr -maxwarn 10000 -n system.ndx
cp minimize_restr.top equilibration.top
#python compute_angles.py -f equilibration.g96 -s production.tpr -d $RANSEED
python restraint_topology.py -n equilibration -p .
# production with restraints
grompp -f production.mdp -c equilibration_restr.g96 -p equilibration_restr.top -o production.tpr -maxwarn 10000 -n system.ndx
mdrun -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
""" % vars()
write_file(os.path.join(complex_path, 'run.sh'), contents)
return
def setup_binding_free_energy(protein_pdb_filename, ligand_filename, work_path, nreplicates_solvent = 1, nreplicates_complex = 1):
"""Set up generalized ensemble simulation to calculate a protein-ligand binding free energy.
ARGUMENTS
protein_pdb_filename (string) - filename of PDB file for the protein, without any missing heavy atoms and with appropriate protonation state already assigned
ligand_filename (string) - filename of some type OpenEye can read (e.g. mol2) describing the ligand with suitable docked coordinates
work_path (string) - directory to place files for binding free energy calculation (created if does not already exist)
OPTIONAL ARGUMENTS
nreplicates_solvent (int) - number of replicates to set up of solvated ligand (default: 1)
nreplicates_complex (int) - number of replicates to set up of the complexed ligand (default: 1)
"""
# parameters
forcefield = 'ffamber96' # forcefield
box_clearance_in_nm = 0.3 # clearance from system to box edges (in nm) # DEBUG set to 1.0
# gromacs filenames
pdb2gmx = 'pdb2gmx_d'
editconf = 'editconf_d'
genbox = 'genbox_d'
grompp = 'grompp_d'
mdrun = 'mdrun_d'
# Create directory for binding free energy calculation if it doesn't exist.
#DEBUG os.makedirs(work_path)
# SET UP PROTEIN TOPOLOGY
print "\nConstructing protein topology..."
# Create .top and .gro files for protein using pdb2gmx and desired forcefield
# look up forcefield index
lines = readFile(os.path.join(os.environ['GMXDATA'], 'top', 'FF.dat'))
for forcefield_index in range(1, len(lines)):
if lines[forcefield_index].find(forcefield) != -1: break
forcefield_index -= 1
# run pdb2gmx
protein_gro_filename = os.path.join(work_path, 'protein.gro')
protein_top_filename = os.path.join(work_path, 'protein.top')
command = 'echo %(forcefield_index)d | %(pdb2gmx)s -f %(protein_pdb_filename)s -o %(protein_gro_filename)s -p %(protein_top_filename)s -ignh -H14' % vars()
print command
#DEBUG output = commands.getoutput(command)
#DEBUG print output
# SET UP LIGAND TOPOLOGY
print "\nConstructing ligand topology..."
# Read the ligand into OEMol.
ligand = readMolecule(ligand_filename, normalize = True)
# Generate GAFF parameters for gromacs for small molecule, using antechamber to generate AM1-BCC charges.
ligand_top_filename = os.path.join(work_path,'ligand.top')
ligand_gro_filename = os.path.join(work_path,'ligand.gro')
#DEBUG parameterizeForGromacs(ligand, topology_filename = ligand_top_filename, coordinate_filename = ligand_gro_filename, charge_model = 'bcc', resname = 'LIG')
# Modify gromacs topology file for alchemical free energy calculation.
#DEBUG perturbGromacsTopology(os.path.join(work_path,'ligand.top'), ligand)
# SET UP SOLVATED LIGAND
print "\nSetting up solvated ligand..."
# Create directory to contain ligand in solution.
solvated_ligand_path = os.path.join(work_path, 'solvated-ligand')
#DEBUG os.makedirs(solvated_ligand_path)
# Convert gromacs topology to an .itp file suitable for inclusion.
top_to_itp(os.path.join(work_path,'ligand.top'), os.path.join(solvated_ligand_path,'solute.itp'), moleculetype = 'solute')
# write enlarged box for solvent
system_gro_filename = os.path.join(solvated_ligand_path, 'system.gro')
system_top_filename = os.path.join(solvated_ligand_path, 'system.top')
command = '%(editconf)s -f %(ligand_gro_filename)s -o %(system_gro_filename)s -bt octahedron -d %(box_clearance_in_nm)f' % vars()
print command
output = commands.getoutput(command)
print output
# write topology for system
contents = """
; amber forcefield
#include "ffamber03.itp"
; my molecule
#include "solute.itp"
; TIP3P water
#include "ffamber_tip3p.itp"
[ system ]
solute
[ molecules ]
; Compound nmols
solute 1
""" % vars()
writeFile(system_top_filename, contents)
# solvate
command = '%(genbox)s -cp %(system_gro_filename)s -cs ffamber_tip3p.gro -o %(system_gro_filename)s -p %(system_top_filename)s' % vars()
print command
output = commands.getoutput(command)
print output
# construct .mdp files
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'output']))
mdpfile.write(os.path.join(solvated_ligand_path, 'minimization-unconstrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'constraints', 'output']))
mdpfile.write(os.path.join(solvated_ligand_path, 'minimization-constrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvated_ligand_path, 'production.mdp'))
# write run script
contents = """\
#/bin/tcsh
source /Users/jchodera/local/gromacs-3.1.4-fe/i386-apple-darwin8.11.1/bin/GMXRC.csh
# unconstrained minimization
%(grompp)s -f minimization-unconstrained.mdp -c system.gro -p system.top -o minimization-unconstrained.tpr -maxwarn 10000
%(mdrun)s -s minimization-unconstrained.tpr -x minimization-unconstrained.xtc -c minimized-unconstrained.gro -e minimization-unconstrained.edr -g minimization-unconstrained.log
# constrained minimization
%(grompp)s -f minimization-constrained.mdp -c minimized-unconstrained.gro -p system.top -o minimization-constrained.tpr -maxwarn 10000
%(mdrun)s -s minimization-constrained.tpr -x minimization-constrained.xtc -c minimized-constrained.gro -e minimization-constrained.edr -g minimization-constrained.log
# production
%(grompp)s -f production.mdp -c minimized-unconstrained.gro -p system.top -o production.tpr -maxwarn 10000
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.gro -e production.edr -g production.log
echo 0 | trjconv_d -f production.xtc -o production.pdb -s production.tpr
""" % vars()
writeFile(os.path.join(solvated_ligand_path, 'run.sh'), contents)
# SET UP SOLVATED COMPLEX
# Create directory to contain complex.
complex_path = os.path.join(work_path, 'solvated-complex')
#DEBUG os.makedirs(complex_path)
# Merge protein and liand .top files.
complex_top_filename = os.path.join(complex_path, 'complex.top')
complex_gro_filename = os.path.join(complex_path, 'complex.gro')
merge_protein_ligand_topologies(protein_top_filename, protein_gro_filename, ligand_top_filename, ligand_gro_filename, complex_top_filename, complex_gro_filename)
# Alter topology file to use ffamber TIP3P.
# system.useTIP3P(complex_top_filename)
# Determine total charge.
total_charge = totalCharge(complex_top_filename)
print "total_charge = %f" % total_charge
# Create a System object to help us out.
system = System(protein_pdb_filename, finalOutputDir = complex_path, finalOutputName = "protein", useff = forcefield)
system.setup.setSaltConditions('MgCl2', 0.010)
system.setup.set_boxType = 'octahedron'
system.setup.set_boxSoluteDistance(box_clearance_in_nm)
system.files.topfile = complex_top_filename
system.totalChargeBeforeIons = total_charge
# Create a periodic boundary conditions box.
command = '%(editconf)s -bt octahedron -f %(complex_gro_filename)s -o %(complex_gro_filename)s -d %(box_clearance_in_nm)f' % vars()
output = commands.getoutput(command)
print output
# solvate the box with TIP3P water
command = '%(genbox)s -cp %(complex_gro_filename)s -cs ffamber_tip3p.gro -o %(complex_gro_filename)s -p %(complex_top_filename)s' % vars()
output = commands.getoutput(command)
print output
# calculate how many ions to add to the box
[np, nn, nwaters] = system.counterions()
print "Adding %(np)d positive and %(nn)d negative to %(nwaters)d waters" % vars()
# minimize the system
### make a tpr file with grompp
#grompp = '%s/grompp -f %s -c %s -o %s -p %s '%(os.environ['GMXPATH'], self.files.mdpfile_Minimization, self.files.grofile, self.files.tprfile, self.files.topfile)
#self.rungmx( grompp, mockrun=self.mockrun, checkForFatalErrors=self.checkForFatalErrors )
### run minimization
#minimize = '%s/mdrun -v -s %s -c %s '%(os.environ['GMXPATH'], self.files.tprfile, self.files.next_gro() )
#self.rungmx( minimize, mockrun=self.mockrun, checkForFatalErrors=self.checkForFatalErrors )
#self.files.increment_gro() # must increment filename for any new gmx file
# do the rest of the preparation steps
#self.postSolvationPreparationSteps()
# construct .mdp files
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'output']))
mdpfile.write(os.path.join(complex_path, 'minimization-unconstrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'minimization', 'nonbonded-solvent', 'constraints', 'output']))
mdpfile.write(os.path.join(complex_path, 'minimization-constrained.mdp'))
mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat', 'barostat', 'free-energy', 'output']))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(complex_path, 'production.mdp'))
# write run script for minimization and production
contents = """\
#/bin/tcsh
source /Users/jchodera/local/gromacs-3.1.4-fe/i386-apple-darwin8.11.1/bin/GMXRC.csh
# unconstrained minimization
%(grompp)s -f minimization-unconstrained.mdp -c system.gro -p system.top -o minimization-unconstrained.tpr -maxwarn 10000
%(mdrun)s -s minimization-unconstrained.tpr -x minimization-unconstrained.xtc -c minimized-unconstrained.gro -e minimization-unconstrained.edr -g minimization-unconstrained.log
# constrained minimization
%(grompp)s -f minimization-constrained.mdp -c minimized-unconstrained.gro -p system.top -o minimization-constrained.tpr -maxwarn 10000
%(mdrun)s -s minimization-constrained.tpr -x minimization-constrained.xtc -c minimized-constrained.gro -e minimization-constrained.edr -g minimization-constrained.log
# production
%(grompp)s -f production.mdp -c minimized-unconstrained.gro -p system.top -o production.tpr -maxwarn 10000
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.gro -e production.edr -g production.log
echo 0 | trjconv_d -f production.xtc -o production.pdb -s production.tpr
""" % vars()
writeFile(os.path.join(complex_path, 'run.sh'), contents)
# Convert gromacs topology to an .itp file suitable for inclusion.
# top_to_itp(os.path.join(molecule_path,'solute.top'), os.path.join(molecule_path,'solute.itp'), moleculetype = molecule_name)
return
def setupHydrationCalculation(solute, nreplicates=1, verbose=True, jobname="hydration"):
"""Set up an absolute alchemical hydration free energy calculation for the given molecule.
ARGUMENTS
solute (OEMol) - the molecule for which hydration free energy is to be computed (with fully explicit hydrogens) in the desired protonation state.
OPTIONAL ARGUMENTS
nreplicates (integer) - the number of replicates to set up (default: 1)
verbose (boolean) - if True, extra debug information will be printed (default: True)
jobname (string) - string to use for job name (default: "hydration")
NOTES
A directory will be created 'molecules/[molecule name]' as obtained from molecule.GetTitle().
"""
# get current directory
current_path = os.getcwd()
# Center the solute molecule.
OECenter(solute)
# get molecule name
solute_name = molecule.GetTitle()
if verbose:
print solute_name
# create molecule path/directory
work_path = os.path.abspath(os.path.join("molecules", solute_name))
os.makedirs(work_path)
# # Write mol2 file for the molecule.
# writeMolecule(solute, os.path.join(solute_path, 'solute.mol2'))
# # Write GAFF parameters for gromacs, using antechamber to generate AM1-BCC charges.
# parameterizeForGromacs(molecule, topology_filename = os.path.join(molecule_path,'solute.top'), coordinate_filename = os.path.join(molecule_path,'solute.gro'), charge_model = 'bcc', resname = 'MOL')
# # Modify gromacs topology file for alchemical free energy calculation.
# perturbGromacsTopology(os.path.join(molecule_path,'solute.top'), molecule)
# # Convert gromacs topology to an .itp file suitable for inclusion.
# top_to_itp(os.path.join(molecule_path,'solute.top'), os.path.join(molecule_path,'solute.itp'), moleculetype = molecule_name)
# get pathnames
mdrun = globals()["mdrun"]
grompp = globals()["grompp"]
editconf = globals()["editconf"]
# SET UP SOLUTE TOPOLOGY
if verbose:
print "\nCONSTRUCTING SOLUTE TOPOLOGY"
# Get formal charge of ligand.
solute_charge = formalCharge(solute)
if verbose:
print "solute formal charge is %d" % solute_charge
# Write molecule with explicit hydrogens to mol2 file.
print "Writing solute mol2 file..."
solute_mol2_filename = os.path.abspath(os.path.join(work_path, "solute.mol2"))
writeMolecule(solute, solute_mol2_filename)
# Set substructure name (which will become residue name).
print "Modifying molecule name..."
modifySubstructureName(solute_mol2_filename, "MOL")
# Run antechamber to assign GAFF atom types.
print "Running antechamber..."
os.chdir(work_path)
gaff_mol2_filename = os.path.join(work_path, "solute.gaff.mol2")
charge_model = "bcc"
command = (
"antechamber -i %(solute_mol2_filename)s -fi mol2 -o solute.gaff.mol2 -fo mol2 -c %(charge_model)s -nc %(solute_charge)d > antechamber.out"
% vars()
)
if verbose:
print command
output = commands.getoutput(command)
if verbose:
print output
os.chdir(current_path)
# Generate frcmod file for additional GAFF parameters.
solute_frcmod_filename = os.path.join(work_path, "frcmod.solute")
command = "parmchk -i %(gaff_mol2_filename)s -f mol2 -o %(solute_frcmod_filename)s" % vars()
if verbose:
print command
output = commands.getoutput(command)
if verbose:
print output
# Run LEaP to generate topology / coordinates.
solute_prmtop_filename = os.path.join(work_path, "solute.prmtop")
solute_crd_filename = os.path.join(work_path, "solute.crd")
solute_off_filename = os.path.join(work_path, "solute.off")
tleap_input_filename = os.path.join(work_path, "setup-solute.leap.in")
tleap_output_filename = os.path.join(work_path, "setup-solute.leap.out")
contents = (
"""
# Load GAFF parameters.
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(solute_frcmod_filename)s
# load solute
solute = loadMol2 %(gaff_mol2_filename)s
# check the solute
check solute
# report net charge
charge solute
# save AMBER parameters
saveAmberParm solute %(solute_prmtop_filename)s %(solute_crd_filename)s
# write .off file
saveOff solute %(solute_off_filename)s
# exit
quit
"""
% vars()
)
write_file(tleap_input_filename, contents)
command = "tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s" % vars()
output = commands.getoutput(command)
# extract total charge
solute_charge = commands.getoutput(
'grep "Total unperturbed charge" %(tleap_output_filename)s | cut -c 27-' % vars()
)
solute_charge = int(round(float(solute_charge))) # round to nearest whole charge
if verbose:
print "solute charge is %d" % solute_charge
# PREPARE SOLVATED SOLUTE
print "\nPREPARING SOLVATED SOLUTE"
# create the directory if it doesn't exist
solvent_path = os.path.join(work_path, "solvent")
if not os.path.exists(solvent_path):
os.makedirs(solvent_path)
# solvate the solute
print "Solvating the solute with tleap..."
system_prmtop_filename = os.path.join(solvent_path, "system.prmtop")
system_crd_filename = os.path.join(solvent_path, "system.crd")
tleap_input_filename = os.path.join(solvent_path, "setup-system.leap.in")
tleap_output_filename = os.path.join(solvent_path, "setup-system.leap.out")
clearance = globals()["clearance"] # clearance around solute (in A)
contents = (
"""
source leaprc.ff99
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(solute_frcmod_filename)s
# Load solute.
loadOff %(solute_off_filename)s
# Create system.
system = combine { solute }
"""
% vars()
)
# add counterions
if solute_charge != 0:
nions = abs(solute_charge)
if solute_charge < 0:
iontype = "Na+"
if solute_charge > 0:
iontype = "Cl-"
contents += (
"""
# Add counterions.
addions system %(iontype)s %(nions)d
"""
% vars()
)
#
contents += (
"""
# Solvate in truncated octahedral box.
solvateBox system TIP3PBOX %(clearance)f iso
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
"""
% vars()
)
write_file(tleap_input_filename, contents)
command = "tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s" % vars()
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv("MMTOOLSPATH"), "converters", "amb2gmx.pl")
system_prefix = os.path.join(solvent_path, "system")
os.chdir(solvent_path)
command = "%(amb2gmx)s --prmtop system.prmtop --crd system.crd --outname system" % vars()
print command
output = commands.getoutput(command)
print output
os.chdir(current_path)
# Extract box size.
g96_filename = os.path.join(solvent_path, "system.g96")
g96_file = open(g96_filename, "r")
g96_lines = g96_file.readlines()
g96_file.close()
box_size = zeros([3], float32)
for line_number in range(len(g96_lines)):
if g96_lines[line_number][0:3] == "BOX":
# parse line with box size in nm
line = g96_lines[line_number + 1]
elements = line.split()
box_size[0] = float(elements[0]) * 10.0
box_size[1] = float(elements[1]) * 10.0
box_size[2] = float(elements[2]) * 10.0
print "box_size = "
print box_size
# make a PDB file for checking
print "Converting system to PDB..."
os.chdir(solvent_path)
command = "cat system.crd | ambpdb -p system.prmtop > system.pdb" % vars()
output = commands.getoutput(command)
print output
os.chdir(current_path)
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(solvent_path, "system.top")
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, "atoms")
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if nelements < 8:
continue
# parse line
atom = dict()
atom["nr"] = int(elements[0])
atom["type"] = elements[1]
atom["resnr"] = int(elements[2])
atom["residue"] = elements[3]
atom["atom"] = elements[4]
atom["cgnr"] = int(elements[5])
atom["charge"] = float(elements[6])
atom["mass"] = float(elements[7])
# add those atoms in the solute to our list
if atom["residue"] == "MOL":
perturb_atom_indices.append(atom["nr"])
perturbGromacsTopology(
system_top_filename,
solute,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices,
)
# set up replicates
for replicate in range(nreplicates):
# Create replicate directory.
working_path = os.path.join(solvent_path, "%d" % replicate)
os.makedirs(working_path)
# TODO: Modify solute coordinates in system.g96 to cycle through available conformations.
# set up mdp files
print "Writing mdp files for replicate %d..." % replicate
mdpfile = MdpFile(compose_blocks(["header", "minimization", "nonbonded-solvent", "constraints", "output"]))
mdpfile.write(os.path.join(working_path, "minimize.mdp"))
mdpfile = MdpFile(
compose_blocks(
["header", "dynamics", "nonbonded-solvent", "constraints", "thermostat", "barostat", "output"]
)
)
mdpfile.setParameter("nsteps", "10000") # 20 ps equilibration
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(working_path, "equilibration.mdp"))
mdpfile = MdpFile(
compose_blocks(
[
"header",
"dynamics",
"nonbonded-solvent",
"constraints",
"thermostat",
"barostat",
"free-energy",
"output",
]
)
)
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(working_path, "production.mdp"))
# write run script
print "Writing run script..."
solvent_path = os.path.abspath(solvent_path)
contents = (
"""\
set#!/bin/tcsh
#BSUB -J %(jobname)s_solvent
#BSUB -o %(working_path)s/outfile.%%J
#BSUB -e %(working_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source $GMXRC
cd %(working_path)s
# constrained minimize
%(grompp)s -f minimize.mdp -c ../system.g96 -p ../system.top -o minimize.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# equilibration
%(grompp)s -f equilibration.mdp -c minimize.g96 -p ../system.top -o equilibration.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.g96 -e equilibration.edr -g equilibration.log
# production
%(grompp)s -f production.mdp -c equilibration.g96 -p ../system.top -o production.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
"""
% vars()
)
write_file(os.path.join(working_path, "run.sh"), contents)
# SET UP VACUUM SIMULATION
# construct pathname for vacuum simulations
vacuum_path = os.path.join(work_path, "vacuum")
if not os.path.exists(vacuum_path):
os.makedirs(vacuum_path)
# solvate the solute
print "Preparing vacuum solute with tleap..."
system_prmtop_filename = os.path.join(vacuum_path, "system.prmtop")
system_crd_filename = os.path.join(vacuum_path, "system.crd")
tleap_input_filename = os.path.join(vacuum_path, "setup-system.leap.in")
tleap_output_filename = os.path.join(vacuum_path, "setup-system.leap.out")
clearance = 50.0 # clearance in A
contents = (
"""
source leaprc.ff99
source leaprc.gaff
# load antechamber-generated additional parameters
mods = loadAmberParams %(solute_frcmod_filename)s
# Load solute.
loadOff %(solute_off_filename)s
# Create system.
system = combine { solute }
"""
% vars()
)
# add counterions
if solute_charge != 0:
nions = abs(solute_charge)
if solute_charge < 0:
iontype = "Na+"
if solute_charge > 0:
iontype = "Cl-"
contents += (
"""
# Add counterions.
addions system %(iontype)s %(nions)d
"""
% vars()
)
#
contents += (
"""
# Create big box.
setBox system centers %(clearance)f iso
# Check the system
check system
# Write the system
saveamberparm system %(system_prmtop_filename)s %(system_crd_filename)s
"""
% vars()
)
write_file(tleap_input_filename, contents)
command = "tleap -f %(tleap_input_filename)s > %(tleap_output_filename)s" % vars()
output = commands.getoutput(command)
# convert to gromacs
print "Converting to gromacs..."
amb2gmx = os.path.join(os.getenv("MMTOOLSPATH"), "converters", "amb2gmx.pl")
os.chdir(vacuum_path)
command = "%(amb2gmx)s --prmtop system.prmtop --crd system.crd --outname system" % vars()
print command
output = commands.getoutput(command)
print output
os.chdir(current_path)
# make a PDB file for checking
print "Converting system to PDB..."
os.chdir(vacuum_path)
command = "cat system.crd | ambpdb -p system.prmtop > system.pdb" % vars()
output = commands.getoutput(command)
print output
os.chdir(current_path)
# write enlarged box for solvent because LEaP doesn't do it right.
system_g96_filename = os.path.join(vacuum_path, "system.g96")
command = "%s -f %s -o %s -bt cubic -box %f %f %f -center 0.0 0.0 0.0" % (
editconf,
system_g96_filename,
system_g96_filename,
box_size[0] / 10.0,
box_size[1] / 10.0,
box_size[2] / 10.0,
)
print command
output = commands.getoutput(command)
print output
# set up perturbation
print "Modifying topology file for perturbation..."
system_top_filename = os.path.join(vacuum_path, "system.top")
lines = read_file(system_top_filename)
perturb_atom_indices = list()
indices = extract_section(lines, "atoms")
nions = 0
for index in indices:
# extract the line
line = stripcomments(lines[index])
# parse the line
elements = line.split()
nelements = len(elements)
# skip if not all elements found
if nelements < 8:
continue
# parse line
atom = dict()
atom["nr"] = int(elements[0])
atom["type"] = elements[1]
atom["resnr"] = int(elements[2])
atom["residue"] = elements[3]
atom["atom"] = elements[4]
atom["cgnr"] = int(elements[5])
atom["charge"] = float(elements[6])
atom["mass"] = float(elements[7])
# DEBUG
# add those atoms in the solute to our list
if atom["residue"] == "MOL":
perturb_atom_indices.append(atom["nr"])
# add counterions to balance solute
if ((atom["residue"] == "Na+") or (atom["residue"] == "Cl-")) and (nions < abs(solute_charge)):
print "WARNING: Solute has net charge of %(solute_charge)d -- will perturb counterions too." % vars()
perturb_atom_indices.append(atom["nr"])
nions += 1
perturbGromacsTopology(
system_top_filename,
solute,
perturb_torsions=True,
perturb_vdw=True,
perturb_charges=True,
perturb_atom_indices=perturb_atom_indices,
)
# construct replicates
for replicate in range(nreplicates):
# Create replicate path.
working_path = os.path.join(vacuum_path, "%d" % replicate)
os.makedirs(working_path)
# TODO: Modify solute coordinates in system.g96 to cycle through available conformations.
# construct .mdp files
print "Writing mdp files for replicate %d..." % replicate
mdpfile = MdpFile(compose_blocks(["header", "minimization", "nonbonded-vacuum", "constraints", "output"]))
cutoff = box_size.min() / 10.0 / 2.0 - 0.0001 # compute cutoff
mdpfile.setParameter("rlist", "%f" % cutoff)
mdpfile.setParameter("rvdw", "%f" % cutoff)
mdpfile.setParameter("rcoulomb", "%f" % cutoff)
mdpfile.write(os.path.join(working_path, "minimize.mdp"))
mdpfile = MdpFile(
compose_blocks(
["header", "dynamics", "nonbonded-vacuum", "constraints", "thermostat", "free-energy", "output"]
)
)
mdpfile.randomizeSeed() # randomize velocities
mdpfile.setParameter("rlist", "%f" % cutoff)
mdpfile.setParameter("rvdw", "%f" % cutoff)
mdpfile.setParameter("rcoulomb", "%f" % cutoff)
mdpfile.write(os.path.join(working_path, "production.mdp"))
# write shell script for minimization and production
contents = (
"""\
#!/bin/tcsh
#BSUB -J %(jobname)s_vacuum
#BSUB -o %(working_path)s/outfile.%%J
#BSUB -e %(working_path)s/errfile.%%J
#BSUB -n 1
#BSUB -M 2000000
#BSUB -W 168:0
source $GMXRC
cd %(working_path)s
# constrained minimize
%(grompp)s -f minimize.mdp -c ../system.g96 -p ../system.top -o minimize.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s minimize.tpr -x minimize.xtc -c minimize.g96 -e minimize.edr -g minimize.log
# production
%(grompp)s -f production.mdp -c minimize.g96 -p ../system.top -o production.tpr -maxwarn 10000 -n ../system.ndx
%(mdrun)s -s production.tpr -o production.trr -x production.xtc -c production.g96 -e production.edr -g production.log
# signal completion
mv run.sh run.sh.done
"""
% vars()
)
write_file(os.path.join(working_path, "run.sh"), contents)
return
def setupHydrationCalculation(molecule):
"""
"""
# get current directory
current_path = os.getcwd()
# Center the molecule.
OECenter(molecule)
# get molecule name
molecule_name = molecule.GetTitle()
print molecule_name
# create molecule path/directory
molecule_path = os.path.abspath(os.path.join('molecules', molecule_name))
os.makedirs(molecule_path)
# Write mol2 file for the molecule.
writeMolecule(molecule, os.path.join(molecule_path, 'solute.mol2'))
# Write GAFF parameters for gromacs, using antechamber to generate AM1-BCC charges.
parameterizeForGromacs(
molecule,
topology_filename=os.path.join(molecule_path, 'solute.top'),
coordinate_filename=os.path.join(molecule_path, 'solute.gro'),
charge_model='bcc',
resname='MOL')
# Modify gromacs topology file for alchemical free energy calculation.
perturbGromacsTopology(os.path.join(molecule_path, 'solute.top'), molecule)
# Convert gromacs topology to an .itp file suitable for inclusion.
top_to_itp(os.path.join(molecule_path, 'solute.top'),
os.path.join(molecule_path, 'solute.itp'),
moleculetype=molecule_name)
# SET UP VACUUM SIMULATION
# construct pathname for vacuum simulations
vacuum_path = os.path.join(molecule_path, 'vacuum')
os.mkdir(vacuum_path)
os.chdir(vacuum_path)
# create molecule in enlarged box
command = 'editconf_d -f ../solute.gro -o system.gro -bt cubic -d 50.0 -center 0.0 0.0 0.0'
print command
output = commands.getoutput(command)
print output
# create topology file
topology_contents = """
; amber forcefield
#include "ffamber03.itp"
; my molecule
#include "../solute.itp"
[ system ]
%(molecule_name)s in vacuum
[ molecules ]
; Compound nmols
%(molecule_name)s 1
""" % vars()
writeFile('system.top', topology_contents)
# construct .mdp files
mdpfile = MdpFile(
compose_blocks(
['header', 'minimization', 'nonbonded-vacuum', 'output']))
mdpfile.write(os.path.join(vacuum_path, 'minimization-unconstrained.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-vacuum', 'constraints',
'output'
]))
mdpfile.write(os.path.join(vacuum_path, 'minimization-constrained.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-vacuum', 'constraints',
'thermostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(vacuum_path, 'production.mdp'))
# write shell script for minimization and production
contents = """\
#/bin/tcsh
source /Users/jchodera/local/gromacs-3.1.4-fe/i386-apple-darwin8.11.1/bin/GMXRC.csh
setenv GROMPP grompp_d
setenv MDRUN mdrun_d
# unconstrained minimization
$GROMPP -f minimization-unconstrained.mdp -c system.gro -p system.top -o minimization-unconstrained.tpr -maxwarn 10000
$MDRUN -s minimization-unconstrained.tpr -x minimization-unconstrained.xtc -c minimized-unconstrained.gro -e minimization-unconstrained.edr -g minimization-unconstrained.log
# constrained minimization
$GROMPP -f minimization-constrained.mdp -c minimized-unconstrained.gro -p system.top -o minimization-constrained.tpr -maxwarn 10000
$MDRUN -s minimization-constrained.tpr -x minimization-constrained.xtc -c minimized-constrained.gro -e minimization-constrained.edr -g minimization-constrained.log
# production
$GROMPP -f production.mdp -c minimized-constrained.gro -p system.top -o production.tpr -maxwarn 10000
$MDRUN -v -s production.tpr -o production.trr -x production.xtc -c production.gro -e production.edr -g production.log
echo 0 | trjconv_d -f production.xtc -o production.pdb -s production.tpr
""" % vars()
writeFile(os.path.join(vacuum_path, 'run.sh'), contents)
# SET UP SOLVATED CALCULATION
# construct pathname for solvent simulations
solvated_path = os.path.join(molecule_path, 'solvent')
os.mkdir(solvated_path)
os.chdir(solvated_path)
# write enlarged box for solvent
command = 'editconf_d -f ../solute.gro -o solute.gro -bt octahedron -d 1.0 -center 0.0 0.0 0.0'
print command
output = commands.getoutput(command)
print output
# write topology
topology_contents = """
; amber forcefield
#include "ffamber03.itp"
; my molecule
#include "../solute.itp"
; TIP3P water
#include "ffamber_tip3p.itp"
[ system ]
%(molecule_name)s in solvent
[ molecules ]
; Compound nmols
%(molecule_name)s 1
""" % vars()
writeFile('system.top', topology_contents)
# solvate
command = 'genbox_d -cp solute.gro -cs ffamber_tip3p.gro -o system.gro -p system.top'
# command = 'genbox_d -cp solute.gro -cs tip3p-pme-waterbox.gro -o system.gro -p system.top' # use our special tip3p box
# command = 'genbox_d -cp solute.gro -cs tip3p-pme-waterbox.gro -o system.gro -p system.top -vdwd 0.0' # insert waters but don't cull overlap
print command
output = commands.getoutput(command)
print output
# construct .mdp files
mdpfile = MdpFile(
compose_blocks(
['header', 'minimization', 'nonbonded-solvent', 'output']))
mdpfile.write(os.path.join(solvated_path,
'minimization-unconstrained.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'minimization', 'nonbonded-solvent', 'constraints',
'output'
]))
mdpfile.write(os.path.join(solvated_path, 'minimization-constrained.mdp'))
# mdpfile = MdpFile(compose_blocks(['header', 'dynamics', 'nonbonded-solvent', 'constraints', 'thermostat-equilibration', 'barostat', 'output']))
# mdpfile.randomizeSeed() # randomize velocities
# mdpfile.setParameter('nsteps', '25000') # 10 ps equilibration
# mdpfile.write(os.path.join(solvated_path, 'equilibration.mdp'))
mdpfile = MdpFile(
compose_blocks([
'header', 'dynamics', 'nonbonded-solvent', 'constraints',
'thermostat', 'barostat', 'free-energy', 'output'
]))
mdpfile.randomizeSeed() # randomize velocities
mdpfile.write(os.path.join(solvated_path, 'production.mdp'))
# write run script
contents = """\
#/bin/tcsh
source /Users/jchodera/local/gromacs-3.1.4-fe/i386-apple-darwin8.11.1/bin/GMXRC.csh
setenv GROMPP grompp_d
setenv MDRUN mdrun_d
# unconstrained minimization
$GROMPP -f minimization-unconstrained.mdp -c system.gro -p system.top -o minimization-unconstrained.tpr -maxwarn 10000
$MDRUN -s minimization-unconstrained.tpr -x minimization-unconstrained.xtc -c minimized-unconstrained.gro -e minimization-unconstrained.edr -g minimization-unconstrained.log
# constrained minimization
$GROMPP -f minimization-constrained.mdp -c minimized-unconstrained.gro -p system.top -o minimization-constrained.tpr -maxwarn 10000
$MDRUN -s minimization-constrained.tpr -x minimization-constrained.xtc -c minimized-constrained.gro -e minimization-constrained.edr -g minimization-constrained.log
# equilibration
#$GROMPP -f equilibration.mdp -c minimized-constrained.gro -p system.top -o equilibration.tpr -maxwarn 10000
#$MDRUN -v -s equilibration.tpr -o equilibration.trr -x equilibration.xtc -c equilibration.gro -e equilibration.edr -g equilibration.log
#echo 0 | trjconv_d -f equilibration.xtc -o equilibration.pdb -s equilibration.tpr
# production
$GROMPP -f production.mdp -c minimized-constrained.gro -p system.top -o production.tpr -maxwarn 10000
#$GROMPP -f production.mdp -c equilibration.gro -p system.top -o production.tpr -maxwarn 10000
#$GROMPP -f production.mdp -c system.gro -p system.top -o production.tpr -maxwarn 10000
$MDRUN -v -s production.tpr -o production.trr -x production.xtc -c production.gro -e production.edr -g production.log
echo 0 | trjconv_d -f production.xtc -o production.pdb -s production.tpr
""" % vars()
writeFile(os.path.join(solvated_path, 'run.sh'), contents)
# restore working directory
os.chdir(current_path)
return