def bilayer_sorter(structure,ndx='system-groups'):
"""
Divide the system into groups.
"""
if 'protein_ready' in wordspace:
gmx('make_ndx',structure=structure,ndx='%s-inspect'%structure,
log='make-ndx-%s-inspect'%structure,inpipe="q\n")
with open(wordspace['step']+'log-make-ndx-%s-inspect'%structure) as fp: lines = fp.readlines()
#---find the protein group because it may not be obvious in CGMD
make_ndx_sifter = '^\s*([0-9]+)\s*Protein'
protein_group = int(re.findall(make_ndx_sifter,
next(i for i in lines if re.match(make_ndx_sifter,i)))[0])
group_selector = "\n".join([
"keep %s"%protein_group,
"name 0 PROTEIN",
#---! hacked
" || ".join(['r '+r for r in wordspace['lipids']+['PIP2']]),
"name 1 LIPIDS",
" || ".join(['r '+r for r in ['W','ION',wordspace['cation'],wordspace['anion']]]),
"name 2 SOLVENT",
"0 | 1 | 2","name 3 SYSTEM","q"])+"\n"
else:
group_selector = "\n".join([
"keep 0",
"name 0 SYSTEM",
" || ".join(['r '+r for r in wordspace['lipids']]),
"name 1 LIPIDS",
" || ".join(['r '+r for r in ['W','ION',wordspace['cation'],wordspace['anion']]]),
"name 2 SOLVENT","q"])+"\n"
gmx('make_ndx',structure='system',ndx=ndx,log='make-ndx-groups',
inpipe=group_selector)
def vacuum_pack(structure='vacuum',
name='vacuum-pack',
gro='vacuum-packed',
pbc='nojump'):
"""
Pack the lipids in the plane, gently.
"""
gmx('grompp',
base='md-%s' % name,
top='vacuum',
structure=structure,
log='grompp-%s' % name,
mdp='input-md-%s-eq-in' % name,
flag='-maxwarn 100')
gmx('mdrun', base='md-%s' % name, log='mdrun-%s' % name, skip=True)
if pbc:
remove_jump(structure='md-%s' % name,
tpr='md-' + name,
gro='md-%s-%s' % (name, pbc))
filecopy(wordspace['step'] + 'md-%s-%s.gro' % (name, pbc),
wordspace['step'] + '%s.gro' % gro)
else:
filecopy(wordspace['step'] + 'md-%s' % gro,
wordspace['step'] + '%s.gro' % gro)
boxdims_old, boxdims = get_box_vectors(gro)
wordspace['bilayer_dimensions_slab'][:2] = boxdims_old[:2]
def get_box_vectors(structure,gro=None,d=0,log='checksize'):
"""
Return the box vectors.
"""
if not gro: gro = structure+'-check-box'
#---note that we consult the command_library here
gmx('editconf',structure=structure,gro=gro,
log='editconf-%s'%log,flag='-d %d'%d)
with open(wordspace['step']+'log-editconf-%s'%log,'r') as fp: lines = fp.readlines()
box_vector_regex = '\s*box vectors\s*\:\s*([^\s]+)\s+([^\s]+)\s+([^\s]+)'
box_vector_new_regex = '\s*new box vectors\s*\:\s*([^\s]+)\s+([^\s]+)\s+([^\s]+)'
runon_regex = '^\s*([-]?[0-9]+\.?[0-9]{0,3})\s*([-]?[0-9]+\.?[0-9]{0,3})\s*([-]?[0-9]+\.?[0-9]{0,3})'
old_line = [l for l in lines if re.match(box_vector_regex,l)][0]
vecs_old = re.findall('\s*box vectors\s*:([^\(]+)',old_line)[0]
try:
#---sometimes the numbers run together
try: vecs_old = [float(i) for i in vecs_old.strip(' ').split()]
except: vecs_old = [float(i) for i in re.findall(runon_regex,vecs_old)[0]]
#---repeat for new box vectors
new_line = [l for l in lines if re.match(box_vector_new_regex,l)][0]
vecs_new = re.findall('\s*box vectors\s*:([^\(]+)',new_line)[0]
try: vecs_new = [float(i) for i in vecs_new.strip(' ').split()]
except: vecs_new = [float(i) for i in re.findall(runon_regex,vecs_new)[0]]
except:
import pdb;pdb.set_trace()
#---no need to keep the output since it is a verbatim copy for diagnostic only
os.remove(wordspace['step']+gro+'.gro')
#import pdb;pdb.set_trace()
return vecs_old,vecs_new
def remove_jump(structure,tpr,gro):
"""
Correct that thing where the bilayer crosses the PBCs and gets split.
"""
gmx('make_ndx',ndx=structure,structure=structure,inpipe="keep 0\nq\n",log='make-ndx-nojump')
gmx('trjconv',ndx=structure,structure=structure,gro=gro,tpr=tpr,
log='trjconv-%s-nojump'%structure,flag='-pbc nojump')
os.remove(wordspace['step']+'log-'+'make-ndx-nojump')
def solvate_bilayer(structure='vacuum'):
"""
Solvate a CGMD bilayer (possibly with proteins) avoiding overlaps.
"""
#---check the size of the slab
incoming_structure = str(structure)
boxdims_old,boxdims = get_box_vectors(structure)
#---! standardize these?
basedim = 3.64428
waterbox = 'inputs/martini-water'
#---make an oversized water box
newdims = boxdims_old[:2]+[wordspace['solvent_thickness']]
#import pdb;pdb.set_trace()
gmx('genconf',structure='martini-water',gro='solvate-empty-uncentered-untrimmed',
nbox=' '.join([str(int(i/basedim+1)) for i in newdims]),log='genconf')
#---trimming waters
with open(wordspace['step']+'solvate-empty-uncentered-untrimmed.gro','r') as fp:
lines = fp.readlines()
modlines = []
for line in lines[2:-1]:
coords = [float(i) for i in line[20:].split()][:3]
if all([coords[i]<newdims[i] for i in range(3)]): modlines.append(line)
with open(wordspace['step']+'solvate-empty-uncentered.gro','w') as fp:
fp.write(lines[0])
fp.write(str(len(modlines))+'\n')
for l in modlines: fp.write(l)
fp.write(lines[-1])
#---update waters
structure='solvate-empty-uncentered'
component('W',count=count_molecules(structure,'W'))
#---translate the water box
gmx('editconf',structure=structure,gro='solvate-water-shifted',
flag='-translate 0 0 %f'%(wordspace['bilayer_dimensions_slab'][2]/2.),log='editconf-solvate-shift')
#---combine and trim with new box vectors
#---! skipping minimization?
structure = 'solvate-water-shifted'
boxdims_old,boxdims = get_box_vectors(structure)
boxvecs = wordspace['bilayer_dimensions_slab'][:2]+[wordspace['bilayer_dimensions_slab'][2]+boxdims[2]]
gro_combinator('%s.gro'%incoming_structure,structure,box=boxvecs,
cwd=wordspace['step'],gro='solvate-dense')
structure = 'solvate-dense'
trim_waters(structure=structure,gro='solvate',boxcut=False,
gap=wordspace['protein_water_gap'],method='cgmd',boxvecs=boxvecs)
structure = 'solvate'
nwaters = count_molecules(structure,'W')
component('W',count=nwaters)
wordspace['bilayer_dimensions_solvate'] = boxvecs
wordspace['water_without_ions'] = nwaters
def equilibrate(groups=None,structure='system'):
"""
equilibrate()
Standard equilibration procedure.
"""
#---sequential equilibration stages
seq = wordspace.equilibration.split(',') if wordspace.equilibration else []
for eqnum,name in enumerate(seq):
if not equilibrate_check(name):
gmx('grompp',base='md-%s'%name,top='system',
structure=structure if eqnum == 0 else 'md-%s'%seq[eqnum-1],
log='grompp-%s'%name,mdp='input-md-%s-eq-in'%name,
flag=('' if not groups else '-n %s'%groups)+' -maxwarn 10')
gmx('mdrun',base='md-%s'%name,log='mdrun-%s'%name,skip=True)
assert os.path.isfile(wordspace['step']+'md-%s.gro'%name)
checkpoint()
#---first part of the equilibration/production run
name = 'md.part0001'
if not equilibrate_check(name) or seq == []:
gmx('grompp',base=name,top='system',
structure='md-%s'%seq[-1] if seq else structure,
log='grompp-0001',mdp='input-md-in',
flag='' if not groups else '-n %s'%groups)
gmx('mdrun',base=name,log='mdrun-0001')
#---we don't assert that the file exists here because the user might kill it and upload
checkpoint()
def minimize(name,method='steep',top=None):
"""
minimize(name,method='steep')
Standard minimization procedure.
"""
gmx('grompp',base='em-%s-%s'%(name,method),top=name if not top else re.sub('^(.+)\.top$',r'\1',top),
structure=name,log='grompp-%s-%s'%(name,method),mdp='input-em-%s-in'%method,skip=True)
assert os.path.isfile(wordspace['step']+'em-%s-%s.tpr'%(name,method))
gmx('mdrun',base='em-%s-%s'%(name,method),log='mdrun-%s-%s'%(name,method))
filecopy(wordspace['step']+'em-'+'%s-%s.gro'%(name,method),
wordspace['step']+'%s-minimized.gro'%name)
checkpoint()
def count_molecules(structure,resname):
"""
Count the number of molecules in a system using make_ndx.
"""
gmx('make_ndx',structure=structure,ndx=structure+'-count',
log='make-ndx-%s-check'%structure,inpipe='q\n')
with open(wordspace['step']+'log-make-ndx-%s-check'%structure) as fp: lines = fp.readlines()
try:
residue_regex = '^\s*[0-9]+\s+%s\s+\:\s+([0-9]+)\s'%resname
count, = [int(re.findall(residue_regex,l)[0]) for l in lines if re.match(residue_regex,l)]
except: raise Exception('cannot find resname "%s" in %s'%(resname,'make-ndx-%s-check'%structure))
return count
def minimize_steep_cg(name):
"""
minimize_steep_cg(name)
Minimization using steepest descent followed by conjugate gradient.
Note that this method has been retired due to reliability issues.
"""
gmx('grompp',
base='em-%s-steep' % name,
top=name,
structure=name,
log='grompp-em-%s-steep' % name,
mdp='input-em-steep-in')
gmx('mdrun',
base='em-%s-steep' % name,
log='mdrun-%s-steep' % name,
skip=True)
#---if first step fails we skip it and try again with the second minimizer
if not os.path.isfile(wordspace['step'] + 'em-%s-steep.gro'):
filecopy(wordspace['step'] + '%s.gro' % name,
wordspace['step'] + 'em-%s-steep.gro' % name)
gmx('grompp',
base='em-%s-cg' % name,
top=name,
structure='em-%s-steep' % name,
log='grompp-%s-cg' % name,
mdp='input-em-cg-in',
skip=True)
gmx('mdrun', base='em-%s-cg' % name, log='mdrun-%s-cg' % name)
select_minimum('%s-steep' % name,
'%s-cg' % name,
gro='%s-minimized' % name)
checkpoint()
def equilibrate(groups=None, structure='system'):
"""
equilibrate()
Standard equilibration procedure.
"""
#---sequential equilibration stages
seq = wordspace.equilibration.split(',') if wordspace.equilibration else []
for eqnum, name in enumerate(seq):
if not equilibrate_check(name):
gmx('grompp',
base='md-%s' % name,
top='system',
structure=structure if eqnum == 0 else 'md-%s' %
seq[eqnum - 1],
log='grompp-%s' % name,
mdp='input-md-%s-eq-in' % name,
flag=('' if not groups else '-n %s' % groups) + ' -maxwarn 10')
gmx('mdrun', base='md-%s' % name, log='mdrun-%s' % name, skip=True)
assert os.path.isfile(wordspace['step'] + 'md-%s.gro' % name)
checkpoint()
#---first part of the equilibration/production run
name = 'md.part0001'
if not equilibrate_check(name) or seq == []:
gmx('grompp',
base=name,
top='system',
structure='md-%s' % seq[-1] if seq else structure,
log='grompp-0001',
mdp='input-md-in',
flag='' if not groups else '-n %s' % groups)
gmx('mdrun', base=name, log='mdrun-0001')
#---we don't assert that the file exists here because the user might kill it and upload
checkpoint()
def vacuum_pack(structure='vacuum',name='vacuum-pack',gro='vacuum-packed'):
"""
Pack the lipids in the plane, gently.
"""
gmx('grompp',base='md-%s'%name,top='vacuum',
structure=structure,log='grompp-%s'%name,mdp='input-md-%s-eq-in'%name,
flag='-maxwarn 100')
gmx('mdrun',base='md-%s'%name,log='mdrun-%s'%name,skip=True)
remove_jump(structure='md-%s'%name,tpr='md-'+name,gro='md-%s-nojump'%name)
filecopy(wordspace['step']+'md-%s-nojump.gro'%name,wordspace['step']+'%s.gro'%gro)
boxdims_old,boxdims = get_box_vectors(gro)
wordspace['bilayer_dimensions_slab'][:2] = boxdims_old[:2]
def bilayer_sorter(structure, ndx='system-groups'):
"""
Divide the system into groups.
"""
if 'protein_ready' in wordspace:
gmx('make_ndx',
structure=structure,
ndx='%s-inspect' % structure,
log='make-ndx-%s-inspect' % structure,
inpipe="q\n")
with open(wordspace['step'] +
'log-make-ndx-%s-inspect' % structure) as fp:
lines = fp.readlines()
#---find the protein group because it may not be obvious in CGMD
make_ndx_sifter = '^\s*([0-9]+)\s*Protein'
protein_group = int(
re.findall(make_ndx_sifter,
next(i for i in lines
if re.match(make_ndx_sifter, i)))[0])
group_selector = "\n".join([
"keep %s" % protein_group,
"name 0 PROTEIN",
#---! hacked
" || ".join(['r ' + r for r in wordspace['lipids'] + ['PIP2']]),
"name 1 LIPIDS",
" || ".join([
'r ' + r
for r in ['W', 'ION', wordspace['cation'], wordspace['anion']]
]),
"name 2 SOLVENT",
"0 | 1 | 2",
"name 3 SYSTEM",
"q"
]) + "\n"
else:
group_selector = "\n".join([
"keep 0", "name 0 SYSTEM", " || ".join([
'r ' + r for r in wordspace['lipids']
]), "name 1 LIPIDS", " || ".join([
'r ' + r
for r in ['W', 'ION', wordspace['cation'], wordspace['anion']]
]), "name 2 SOLVENT", "q"
]) + "\n"
gmx('make_ndx',
structure='system',
ndx=ndx,
log='make-ndx-groups',
inpipe=group_selector)
def solvate_bilayer(structure='vacuum'):
"""
Solvate a CGMD bilayer (possibly with proteins) avoiding overlaps.
"""
#---check the size of the slab
incoming_structure = str(structure)
boxdims_old,boxdims = get_box_vectors(structure)
#---check the size of the water box
waterbox = wordspace.water_box
basedim,_ = get_box_vectors(waterbox)
if not all([i==basedim[0] for i in basedim]):
raise Exception('[ERROR] expecting water box "" to be cubic')
else: basedim = basedim[0]
#---make an oversized water box
newdims = boxdims_old[:2]+[wordspace['solvent_thickness']]
gmx('genconf',structure=waterbox,gro='solvate-empty-uncentered-untrimmed',
nbox=' '.join([str(int(i/basedim+1)) for i in newdims]),log='genconf')
#---trim the blank water box
trim_waters(structure='solvate-empty-uncentered-untrimmed',
gro='solvate-empty-uncentered',boxcut=True,boxvecs=newdims,
gap=0.0,method=wordspace.atom_resolution)
#---update waters
structure='solvate-empty-uncentered'
component(wordspace.sol,count=count_molecules(structure,wordspace.sol))
#---translate the water box
gmx('editconf',structure=structure,gro='solvate-water-shifted',
flag='-translate 0 0 %f'%(wordspace['bilayer_dimensions_slab'][2]/2.),log='editconf-solvate-shift')
#---combine and trim with new box vectors
structure = 'solvate-water-shifted'
boxdims_old,boxdims = get_box_vectors(structure)
boxvecs = wordspace['bilayer_dimensions_slab'][:2]+[wordspace['bilayer_dimensions_slab'][2]+boxdims[2]]
gro_combinator('%s.gro'%incoming_structure,structure,box=boxvecs,
cwd=wordspace['step'],gro='solvate-dense')
structure = 'solvate-dense'
#---trim everything so that waters are positioned in the box without steric clashes
trim_waters(structure=structure,gro='solvate',boxcut=False,
gap=wordspace['protein_water_gap'],method=wordspace.atom_resolution,boxvecs=boxvecs)
structure = 'solvate'
nwaters = count_molecules(structure,wordspace.sol)/({'aamd':3.0,'cgmd':1.0}[wordspace.atom_resolution])
if round(nwaters)!=nwaters: raise Exception('[ERROR] fractional water molecules')
else: nwaters = int(nwaters)
component(wordspace.sol,count=nwaters)
wordspace['bilayer_dimensions_solvate'] = boxvecs
wordspace['water_without_ions'] = nwaters
def remove_jump(structure, tpr, gro):
"""
Correct that thing where the bilayer crosses the PBCs and gets split.
"""
gmx('make_ndx',
ndx=structure,
structure=structure,
inpipe="keep 0\nq\n",
log='make-ndx-nojump')
gmx('trjconv',
ndx=structure,
structure=structure,
gro=gro,
tpr=tpr,
log='trjconv-%s-nojump' % structure,
flag='-pbc nojump')
os.remove(wordspace['step'] + 'log-' + 'make-ndx-nojump')
def minimize(name, method='steep', top=None):
"""
minimize(name,method='steep')
Standard minimization procedure.
"""
gmx('grompp',
base='em-%s-%s' % (name, method),
top=name if not top else re.sub('^(.+)\.top$', r'\1', top),
structure=name,
log='grompp-%s-%s' % (name, method),
mdp='input-em-%s-in' % method,
skip=True)
assert os.path.isfile(wordspace['step'] + 'em-%s-%s.tpr' % (name, method))
gmx('mdrun',
base='em-%s-%s' % (name, method),
log='mdrun-%s-%s' % (name, method))
filecopy(wordspace['step'] + 'em-' + '%s-%s.gro' % (name, method),
wordspace['step'] + '%s-minimized.gro' % name)
checkpoint()
def bilayer_middle(structure, gro):
"""
Move the bilayer to the middle of the z-coordinate of the box.
Note that the protein adhesion procedure works best on a slab that is centered on z=0.
This means that the bilayer will be broken across z=0.
For visualization it is better to center it.
"""
gmx('make_ndx',
ndx='system-dry',
structure='counterions-minimized',
inpipe="keep 0\nr %s || r ION || r %s || r %s\n!1\ndel 1\nq\n" %
(wordspace['sol'], wordspace['anion'], wordspace['cation']),
log='make-ndx-center')
#---bilayer slab is near z=0 so it is likely split so we shift by half of the box vector
gmx('trjconv',
structure='counterions-minimized',
gro='counterions-shifted',
ndx='system-dry',
flag='-trans 0 0 %f -pbc mol' %
(wordspace['bilayer_dimensions_solvate'][2] / 2.),
tpr='em-counterions-steep',
log='trjconv-shift',
inpipe="0\n")
#---center everything
gmx('trjconv',
structure='counterions-shifted',
gro='system',
ndx='system-dry',
tpr='em-counterions-steep',
log='trjconv-middle',
inpipe="1\n0\n",
flag='-center -pbc mol')
def get_box_vectors(structure, gro=None, d=0, log='checksize'):
"""
Return the box vectors.
"""
if not gro: gro = structure + '-check-box'
#---note that we consult the command_library here
gmx('editconf',
structure=structure,
gro=gro,
log='editconf-%s' % log,
flag='-d %d' % d)
with open(wordspace['step'] + 'log-editconf-%s' % log, 'r') as fp:
lines = fp.readlines()
box_vector_regex = '\s*box vectors\s*\:\s*([^\s]+)\s+([^\s]+)\s+([^\s]+)'
box_vector_new_regex = '\s*new box vectors\s*\:\s*([^\s]+)\s+([^\s]+)\s+([^\s]+)'
runon_regex = '^\s*([-]?[0-9]+\.?[0-9]{0,3})\s*([-]?[0-9]+\.?[0-9]{0,3})\s*([-]?[0-9]+\.?[0-9]{0,3})'
old_line = [l for l in lines if re.match(box_vector_regex, l)][0]
vecs_old = re.findall('\s*box vectors\s*:([^\(]+)', old_line)[0]
try:
#---sometimes the numbers run together
try:
vecs_old = [float(i) for i in vecs_old.strip(' ').split()]
except:
vecs_old = [float(i) for i in re.findall(runon_regex, vecs_old)[0]]
#---repeat for new box vectors
new_line = [l for l in lines if re.match(box_vector_new_regex, l)][0]
vecs_new = re.findall('\s*box vectors\s*:([^\(]+)', new_line)[0]
try:
vecs_new = [float(i) for i in vecs_new.strip(' ').split()]
except:
vecs_new = [float(i) for i in re.findall(runon_regex, vecs_new)[0]]
except:
import pdb
pdb.set_trace()
#---no need to keep the output since it is a verbatim copy for diagnostic only
os.remove(wordspace['step'] + gro + '.gro')
#import pdb;pdb.set_trace()
return vecs_old, vecs_new
def count_molecules(structure, resname):
"""
Count the number of molecules in a system using make_ndx.
"""
gmx('make_ndx',
structure=structure,
ndx=structure + '-count',
log='make-ndx-%s-check' % structure,
inpipe='q\n')
with open(wordspace['step'] + 'log-make-ndx-%s-check' % structure) as fp:
lines = fp.readlines()
try:
residue_regex = '^\s*[0-9]+\s+%s\s+\:\s+([0-9]+)\s' % resname
count, = [
int(re.findall(residue_regex, l)[0]) for l in lines
if re.match(residue_regex, l)
]
except:
raise Exception('cannot find resname "%s" in %s' %
(resname, 'make-ndx-%s-check' % structure))
return count
def minimize_steep_cg(name):
"""
minimize_steep_cg(name)
Minimization using steepest descent followed by conjugate gradient.
Note that this method has been retired due to reliability issues.
"""
gmx('grompp',base='em-%s-steep'%name,top=name,structure=name,
log='grompp-em-%s-steep'%name,mdp='input-em-steep-in')
gmx('mdrun',base='em-%s-steep'%name,log='mdrun-%s-steep'%name,skip=True)
#---if first step fails we skip it and try again with the second minimizer
if not os.path.isfile(wordspace['step']+'em-%s-steep.gro'):
filecopy(wordspace['step']+'%s.gro'%name,wordspace['step']+'em-%s-steep.gro'%name)
gmx('grompp',base='em-%s-cg'%name,top=name,structure='em-%s-steep'%name,
log='grompp-%s-cg'%name,mdp='input-em-cg-in',skip=True)
gmx('mdrun',base='em-%s-cg'%name,log='mdrun-%s-cg'%name)
select_minimum('%s-steep'%name,'%s-cg'%name,gro='%s-minimized'%name)
checkpoint()
def solvate(structure,top):
"""
solvate(structure,top)
Standard solvate procedure for atomistic protein in water.
"""
#---purge the wordspace of solvent and anions in case we are resuming
for key in [wordspace['anion'],wordspace['cation'],'SOL']:
if key in zip(*wordspace['composition'])[0]:
del wordspace['composition'][zip(*wordspace['composition'])[0].index(key)]
gmx('editconf',structure=structure,gro='solvate-box-alone',
log='editconf-checksize',flag='-d 0')
with open(wordspace['step']+'log-editconf-checksize','r') as fp: lines = fp.readlines()
boxdims = map(lambda y:float(y),re.findall('\s*box vectors \:\s*([^\s]+)\s+([^\s]+)\s+([^\s]+)',
filter(lambda x:re.match('\s*box vectors',x),lines).pop()).pop())
boxvecs = tuple([i+2*wordspace['water_buffer'] for i in boxdims])
center = tuple([i/2. for i in boxvecs])
#---cube is not implemented yet
gmx('editconf',structure=structure,gro='solvate-protein',
flags='-center %f %f %f'%center+' '+'-box %f %f %f'%boxvecs,
log='editconf-center-protein')
gmx('genbox',structure='solvate-protein',solvent=wordspace['solvent'],
gro='solvate-dense',#top='solvate-standard',
log='genbox-solvate')
#---trim waters if the protein_water_gap setting is not False
if 'protein_water_gap' in wordspace and wordspace['protein_water_gap'] != False:
trim_waters(structure='solvate-dense',gro='solvate',
gap=wordspace['protein_water_gap'],boxvecs=boxvecs)
else: filecopy(wordspace['step']+'solvate-dense.gro',wordspace['step']+'solvate.gro')
gmx('make_ndx',structure='solvate',ndx='solvate-water-check',inpipe='q\n',
log='make-ndx-solvate-check')
with open(wordspace['step']+'log-make-ndx-solvate-check','r') as fp: lines = fp.readlines()
nwaters = int(re.findall('\s*[0-9]+\s+Water\s+:\s+([0-9]+)\s+atoms',
filter(lambda x:re.match('\s*[0-9]+\s+Water',x),lines).pop()).pop())/3
wordspace['water_without_ions'] = nwaters
component('SOL',count=nwaters)
#---add the suffix so that water is referred to by its name in the settings
include(wordspace['water'],ff=True)
write_top('solvate.top')
def bilayer_middle(structure,gro):
"""
Move the bilayer to the middle of the z-coordinate of the box.
Note that the protein adhesion procedure works best on a slab that is centered on z=0.
This means that the bilayer will be broken across z=0.
For visualization it is better to center it.
"""
gmx('make_ndx',ndx='system-dry',structure='counterions-minimized',
inpipe="keep 0\nr %s || r ION || r %s || r %s\n!1\ndel 1\nq\n"%(
wordspace['sol'],wordspace['anion'],wordspace['cation']),
log='make-ndx-center')
#---bilayer slab is near z=0 so it is likely split so we shift by half of the box vector
gmx('trjconv',structure='counterions-minimized',gro='counterions-shifted',ndx='system-dry',
flag='-trans 0 0 %f -pbc mol'%(wordspace['bilayer_dimensions_solvate'][2]/2.),
tpr='em-counterions-steep',log='trjconv-shift',inpipe="0\n")
#---center everything
gmx('trjconv',structure='counterions-shifted',gro='system',ndx='system-dry',
tpr='em-counterions-steep',log='trjconv-middle',inpipe="1\n0\n",flag='-center -pbc mol')
def counterions(structure,top,includes=None,ff_includes=None,gro='counterions'):
"""
counterions(structure,top)
Standard procedure for adding counterions.
The resname must be understandable by "r RESNAME" in make_ndx and writes to the top file.
"""
#---we store the water resname in the wordspace as "sol"
resname = wordspace.get('sol','SOL')
#---clean up the composition in case this is a restart
for key in ['cation','anion',resname]:
try: wordspace['composition'].pop(zip(*wordspace['composition'])[0].index(wordspace[key]))
except: pass
component(resname,count=wordspace['water_without_ions'])
#---write the topology file as of the solvate step instead of copying them (genion overwrites top)
write_top('counterions.top')
gmx('grompp',base='genion',structure=structure,
top='counterions',mdp='input-em-steep-in',
log='grompp-genion')
gmx('make_ndx',structure=structure,ndx='solvate-waters',
inpipe='keep 0\nr %s\nkeep 1\nq\n'%resname,
log='make-ndx-counterions-check')
gmx('genion',base='genion',gro=gro,ndx='solvate-waters',
cation=wordspace['cation'],anion=wordspace['anion'],
flag='-conc %f -neutral'%wordspace['ionic_strength'],
log='genion')
with open(wordspace['step']+'log-genion','r') as fp: lines = fp.readlines()
declare_ions = filter(lambda x:re.search('Will try',x)!=None,lines).pop()
ion_counts = re.findall(
'^Will try to add ([0-9]+)\+?\-? ([\w\+\-]+) ions and ([0-9]+) ([\w\+\-]+) ions',
declare_ions).pop()
for ii in range(2): component(ion_counts[2*ii+1],count=ion_counts[2*ii])
component(resname,count=component(resname)-component(ion_counts[1])-component(ion_counts[3]))
if includes:
if type(includes)==str: includes = [includes]
for i in includes: include(i)
if ff_includes:
if type(ff_includes)==str: ff_includes = [ff_includes]
for i in ff_includes: include(i,ff=True)
write_top('counterions.top')
def write_structure_pdb(structure, pdb):
"""
write_structure_pdb(structure,pdb)
Infer the starting residue from the original PDB and write structure.pdb with the correct indices
according to the latest GRO structure (typically counterions.gro).
"""
#---automatically center the protein in the box here and write the final structure
gmx(
'make_ndx',
structure='counterions',
ndx='counterions-groups',
log='make-ndx-counterions',
inpipe='q\n',
)
with open(wordspace['step'] + 'log-make-ndx-counterions', 'r') as fp:
lines = fp.readlines()
relevant = [
filter(lambda x: re.match('\s*[0-9]+\s+%s' % name, x), lines)
for name in ['System', 'Protein']
]
groupdict = dict([
(j[1], int(j[0])) for j in
[re.findall('^\s*([0-9]+)\s(\w+)', x[0])[0] for x in relevant]
])
gmx('trjconv',
ndx='counterions-groups',
structure='counterions-minimized',
inpipe='%d\n%d\n' % (groupdict['Protein'], groupdict['System']),
log='trjconv-counterions-center',
tpr='em-counterions-steep',
gro='system')
with open(wordspace['step'] + pdb, 'r') as fp:
lines = fp.readlines()
startres = int([line for line in lines
if re.match('^ATOM', line)][0][23:26 + 1])
gmx('editconf',
structure=structure,
flag='-o structure.pdb -resnr %d' % startres,
log='editconf-structure-pdb')
def write_structure_pdb(structure,pdb):
"""
write_structure_pdb(structure,pdb)
Infer the starting residue from the original PDB and write structure.pdb with the correct indices
according to the latest GRO structure (typically counterions.gro).
"""
#---automatically center the protein in the box here and write the final structure
gmx('make_ndx',structure='counterions',ndx='counterions-groups',
log='make-ndx-counterions',inpipe='q\n',)
with open(wordspace['step']+'log-make-ndx-counterions','r') as fp: lines = fp.readlines()
relevant = [filter(lambda x:re.match('\s*[0-9]+\s+%s'%name,x),lines) for name in ['System','Protein']]
groupdict = dict([(j[1],int(j[0])) for j in
[re.findall('^\s*([0-9]+)\s(\w+)',x[0])[0] for x in relevant]])
gmx('trjconv',ndx='counterions-groups',structure='counterions-minimized',
inpipe='%d\n%d\n'%(groupdict['Protein'],groupdict['System']),
log='trjconv-counterions-center',tpr='em-counterions-steep',gro='system')
with open(wordspace['step']+pdb,'r') as fp: lines = fp.readlines()
startres = int([line for line in lines if re.match('^ATOM',line)][0][23:26+1])
gmx('editconf',structure=structure,
flag='-o structure.pdb -resnr %d'%startres,
log='editconf-structure-pdb')
def counterions(structure,
top,
includes=None,
ff_includes=None,
gro='counterions'):
"""
counterions(structure,top)
Standard procedure for adding counterions.
The resname must be understandable by "r RESNAME" in make_ndx and writes to the top file.
"""
#---we store the water resname in the wordspace as "sol"
resname = wordspace.get('sol', 'SOL')
#---clean up the composition in case this is a restart
for key in ['cation', 'anion', resname]:
try:
wordspace['composition'].pop(
zip(*wordspace['composition'])[0].index(wordspace[key]))
except:
pass
component(resname, count=wordspace['water_without_ions'])
#---write the topology file as of the solvate step instead of copying them (genion overwrites top)
write_top('counterions.top')
gmx('grompp',
base='genion',
structure=structure,
top='counterions',
mdp='input-em-steep-in',
log='grompp-genion')
gmx('make_ndx',
structure=structure,
ndx='solvate-waters',
inpipe='keep 0\nr %s\nkeep 1\nq\n' % resname,
log='make-ndx-counterions-check')
gmx('genion',
base='genion',
gro=gro,
ndx='solvate-waters',
cation=wordspace['cation'],
anion=wordspace['anion'],
flag='-conc %f -neutral' % wordspace['ionic_strength'],
log='genion')
with open(wordspace['step'] + 'log-genion', 'r') as fp:
lines = fp.readlines()
declare_ions = filter(lambda x: re.search('Will try', x) != None,
lines).pop()
ion_counts = re.findall(
'^Will try to add ([0-9]+)\+?\-? ([\w\+\-]+) ions and ([0-9]+) ([\w\+\-]+) ions',
declare_ions).pop()
for ii in range(2):
component(ion_counts[2 * ii + 1], count=ion_counts[2 * ii])
component(resname,
count=component(resname) - component(ion_counts[1]) -
component(ion_counts[3]))
if includes:
if type(includes) == str: includes = [includes]
for i in includes:
include(i)
if ff_includes:
if type(ff_includes) == str: ff_includes = [ff_includes]
for i in ff_includes:
include(i, ff=True)
write_top('counterions.top')
def multiply(nx=1,ny=1,nz=1,quirky_ions=True):
"""
Make a copy of a simulation box in multiple directions.
"""
factor = nx*ny*nz
#---update the composition
#---if the last step doesn't have the composition we step backwards and pick up requirements
#---note that "protein_ready" is important for the bilayer_sorter
for prereq in ['composition','lipids','cation','anion','protein_ready']:
if prereq not in wordspace:
steplist = detect_last(steplist=True)[::-1]
#---walk backwards through steps until we find the commposition
for ii,i in enumerate(steplist):
oldspace = resume(read_only=True,step=int(re.match('s([0-9]+)-',i).group(1)))
if prereq in oldspace:
wordspace[prereq] = deepcopy(oldspace[prereq])
break
#---if composition is available we continue
wordspace['new_composition'] = [[name,count*factor] for name,count in wordspace['composition']]
kwargs = {}
if 'buffer' in wordspace: kwargs['flag'] = ' -dist %.2f %.2f %.2f'%tuple(wordspace['buffer'])
gmx('genconf',structure='system-input',gro='system-multiply',
nbox="%d %d %d"%(nx,ny,nz),log='genconf-multiply',**kwargs)
#---copy ITP files
for itp in wordspace.itp:
filecopy(wordspace.last_step+itp,wordspace.step+itp)
#---reorder the GRO for convenience
with open(wordspace['step']+'system-multiply.gro') as fp: lines = fp.readlines()
#---collect all unique resiue/atom combinations
combos = list(set([l[5:15] for l in lines]))
#---for each element in the composition, extract all of the residues for that element
lines_reorder = []
lines_reorder.extend(lines[:2])
#---develop a list of filtering rules
keylist = {}
for key,count in wordspace['new_composition']:
if key in [wordspace[i] for i in ['anion','cation']]:
keylist[key] = 'regex',(('ION',key),slice(5,15),'\s*%s\s*%s\s*')
elif re.match('^(p|P)rotein',key) and key+'.itp' in wordspace.itp:
#---custom procedure for finding proteins which have variegated residue numbers
itp = read_itp(wordspace.step+key+'.itp')
residues_starts = []
seq = list(zip(*itp['atoms'])[3])
residues = [i[5:10].strip() for i in lines]
for i in range(len(residues)-len(seq)):
#---minor speed up by checking the first one
if seq[0]==residues[i] and residues[i:i+len(seq)]==seq:
residues_starts.append(i)
keylist[key] = 'slices',[slice(i,i+len(seq)) for i in residues_starts]
else: keylist[key] = 'regex',(key,slice(5,10),'\s*%s\s*')
for key,count in wordspace['new_composition']:
method,details = keylist[key]
if method == 'regex':
key,sl,regex = details
lines_reorder.extend([l for l in lines[2:-1] if re.match(regex%key,l[sl])])
elif method == 'slices':
for sl in details: lines_reorder.extend(lines[sl])
else: raise
lines_reorder.extend([lines[-1]])
with open(wordspace['step']+'system-multiply-reorder.gro','w') as fp:
for line in lines_reorder: fp.write(line)
filecopy(wordspace['step']+'system-multiply-reorder.gro',wordspace['step']+'system.gro')
wordspace['composition'] = tuple(wordspace['new_composition'])
del wordspace['new_composition']
def solvate(structure, top):
"""
solvate(structure,top)
Standard solvate procedure for atomistic protein in water.
"""
#---purge the wordspace of solvent and anions in case we are resuming
for key in [wordspace['anion'], wordspace['cation'], 'SOL']:
if key in zip(*wordspace['composition'])[0]:
del wordspace['composition'][zip(
*wordspace['composition'])[0].index(key)]
gmx('editconf',
structure=structure,
gro='solvate-box-alone',
log='editconf-checksize',
flag='-d 0')
with open(wordspace['step'] + 'log-editconf-checksize', 'r') as fp:
lines = fp.readlines()
boxdims = map(
lambda y: float(y),
re.findall(
'\s*box vectors \:\s*([^\s]+)\s+([^\s]+)\s+([^\s]+)',
filter(lambda x: re.match('\s*box vectors', x),
lines).pop()).pop())
boxvecs = tuple([i + 2 * wordspace['water_buffer'] for i in boxdims])
center = tuple([i / 2. for i in boxvecs])
#---cube is not implemented yet
gmx('editconf',
structure=structure,
gro='solvate-protein',
flags='-center %f %f %f' % center + ' ' + '-box %f %f %f' % boxvecs,
log='editconf-center-protein')
gmx(
'genbox',
structure='solvate-protein',
solvent=wordspace['solvent'],
gro='solvate-dense', #top='solvate-standard',
log='genbox-solvate')
#---trim waters if the protein_water_gap setting is not False
if 'protein_water_gap' in wordspace and wordspace[
'protein_water_gap'] != False:
trim_waters(structure='solvate-dense',
gro='solvate',
gap=wordspace['protein_water_gap'],
boxvecs=boxvecs)
else:
filecopy(wordspace['step'] + 'solvate-dense.gro',
wordspace['step'] + 'solvate.gro')
gmx('make_ndx',
structure='solvate',
ndx='solvate-water-check',
inpipe='q\n',
log='make-ndx-solvate-check')
with open(wordspace['step'] + 'log-make-ndx-solvate-check', 'r') as fp:
lines = fp.readlines()
nwaters = int(
re.findall(
'\s*[0-9]+\s+Water\s+:\s+([0-9]+)\s+atoms',
filter(lambda x: re.match('\s*[0-9]+\s+Water', x),
lines).pop()).pop()) / 3
wordspace['water_without_ions'] = nwaters
component('SOL', count=nwaters)
#---add the suffix so that water is referred to by its name in the settings
include(wordspace['water'], ff=True)
write_top('solvate.top')
def solvate_bilayer(structure='vacuum'):
"""
Solvate a CGMD bilayer (possibly with proteins) avoiding overlaps.
"""
#---check the size of the slab
incoming_structure = str(structure)
boxdims_old, boxdims = get_box_vectors(structure)
#---check the size of the water box
waterbox = wordspace.water_box
basedim, _ = get_box_vectors(waterbox)
if not all([i == basedim[0] for i in basedim]):
raise Exception('[ERROR] expecting water box "" to be cubic')
else:
basedim = basedim[0]
#---make an oversized water box
newdims = boxdims_old[:2] + [wordspace['solvent_thickness']]
gmx('genconf',
structure=waterbox,
gro='solvate-empty-uncentered-untrimmed',
nbox=' '.join([str(int(i / basedim + 1)) for i in newdims]),
log='genconf')
#---trim the blank water box
trim_waters(structure='solvate-empty-uncentered-untrimmed',
gro='solvate-empty-uncentered',
boxcut=True,
boxvecs=newdims,
gap=0.0,
method=wordspace.atom_resolution)
#---update waters
structure = 'solvate-empty-uncentered'
component(wordspace.sol, count=count_molecules(structure, wordspace.sol))
#---translate the water box
gmx('editconf',
structure=structure,
gro='solvate-water-shifted',
flag='-translate 0 0 %f' %
(wordspace['bilayer_dimensions_slab'][2] / 2.),
log='editconf-solvate-shift')
#---combine and trim with new box vectors
structure = 'solvate-water-shifted'
boxdims_old, boxdims = get_box_vectors(structure)
boxvecs = wordspace['bilayer_dimensions_slab'][:2] + [
wordspace['bilayer_dimensions_slab'][2] + boxdims[2]
]
gro_combinator('%s.gro' % incoming_structure,
structure,
box=boxvecs,
cwd=wordspace['step'],
gro='solvate-dense')
structure = 'solvate-dense'
#---trim everything so that waters are positioned in the box without steric clashes
trim_waters(structure=structure,
gro='solvate',
boxcut=False,
gap=wordspace['protein_water_gap'],
method=wordspace.atom_resolution,
boxvecs=boxvecs)
structure = 'solvate'
nwaters = count_molecules(structure, wordspace.sol) / ({
'aamd': 3.0,
'cgmd': 1.0
}[wordspace.atom_resolution])
if round(nwaters) != nwaters:
raise Exception('[ERROR] fractional water molecules')
else:
nwaters = int(nwaters)
component(wordspace.sol, count=nwaters)
wordspace['bilayer_dimensions_solvate'] = boxvecs
wordspace['water_without_ions'] = nwaters
def multiply(nx=1, ny=1, nz=1, quirky_ions=True):
"""
Make a copy of a simulation box in multiple directions.
"""
factor = nx * ny * nz
#---update the composition
#---if the last step doesn't have the composition we step backwards and pick up requirements
#---note that "protein_ready" is important for the bilayer_sorter
for prereq in [
'composition', 'lipids', 'cation', 'anion', 'protein_ready'
]:
if prereq not in wordspace:
steplist = detect_last(steplist=True)[::-1]
#---walk backwards through steps until we find the commposition
for ii, i in enumerate(steplist):
oldspace = resume(read_only=True,
step=int(re.match('s([0-9]+)-', i).group(1)))
if prereq in oldspace:
wordspace[prereq] = deepcopy(oldspace[prereq])
break
#---if composition is available we continue
wordspace['new_composition'] = [[name, count * factor]
for name, count in wordspace['composition']
]
kwargs = {}
if 'buffer' in wordspace:
kwargs['flag'] = ' -dist %.2f %.2f %.2f' % tuple(wordspace['buffer'])
gmx('genconf',
structure='system-input',
gro='system-multiply',
nbox="%d %d %d" % (nx, ny, nz),
log='genconf-multiply',
**kwargs)
#---copy ITP files
for itp in wordspace.itp:
filecopy(wordspace.last_step + itp, wordspace.step + itp)
#---reorder the GRO for convenience
with open(wordspace['step'] + 'system-multiply.gro') as fp:
lines = fp.readlines()
#---collect all unique resiue/atom combinations
combos = list(set([l[5:15] for l in lines]))
#---for each element in the composition, extract all of the residues for that element
lines_reorder = []
lines_reorder.extend(lines[:2])
#---develop a list of filtering rules
keylist = {}
for key, count in wordspace['new_composition']:
if key in [wordspace[i] for i in ['anion', 'cation']]:
keylist[key] = 'regex', (('ION', key), slice(5,
15), '\s*%s\s*%s\s*')
elif re.match('^(p|P)rotein', key) and key + '.itp' in wordspace.itp:
#---custom procedure for finding proteins which have variegated residue numbers
itp = read_itp(wordspace.step + key + '.itp')
residues_starts = []
seq = list(zip(*itp['atoms'])[3])
residues = [i[5:10].strip() for i in lines]
for i in range(len(residues) - len(seq)):
#---minor speed up by checking the first one
if seq[0] == residues[i] and residues[i:i + len(seq)] == seq:
residues_starts.append(i)
keylist[key] = 'slices', [
slice(i, i + len(seq)) for i in residues_starts
]
else:
keylist[key] = 'regex', (key, slice(5, 10), '\s*%s\s*')
for key, count in wordspace['new_composition']:
method, details = keylist[key]
if method == 'regex':
key, sl, regex = details
lines_reorder.extend(
[l for l in lines[2:-1] if re.match(regex % key, l[sl])])
elif method == 'slices':
for sl in details:
lines_reorder.extend(lines[sl])
else:
raise
lines_reorder.extend([lines[-1]])
with open(wordspace['step'] + 'system-multiply-reorder.gro', 'w') as fp:
for line in lines_reorder:
fp.write(line)
filecopy(wordspace['step'] + 'system-multiply-reorder.gro',
wordspace['step'] + 'system.gro')
wordspace['composition'] = tuple(wordspace['new_composition'])
del wordspace['new_composition']
def solvate_bilayer(structure='vacuum'):
"""
Solvate a CGMD bilayer (possibly with proteins) avoiding overlaps.
"""
#---check the size of the slab
incoming_structure = str(structure)
boxdims_old, boxdims = get_box_vectors(structure)
#---! standardize these?
basedim = 3.64428
waterbox = 'inputs/martini-water'
#---make an oversized water box
newdims = boxdims_old[:2] + [wordspace['solvent_thickness']]
#import pdb;pdb.set_trace()
gmx('genconf',
structure='martini-water',
gro='solvate-empty-uncentered-untrimmed',
nbox=' '.join([str(int(i / basedim + 1)) for i in newdims]),
log='genconf')
#---trimming waters
with open(wordspace['step'] + 'solvate-empty-uncentered-untrimmed.gro',
'r') as fp:
lines = fp.readlines()
modlines = []
for line in lines[2:-1]:
coords = [float(i) for i in line[20:].split()][:3]
if all([coords[i] < newdims[i] for i in range(3)]):
modlines.append(line)
with open(wordspace['step'] + 'solvate-empty-uncentered.gro', 'w') as fp:
fp.write(lines[0])
fp.write(str(len(modlines)) + '\n')
for l in modlines:
fp.write(l)
fp.write(lines[-1])
#---update waters
structure = 'solvate-empty-uncentered'
component('W', count=count_molecules(structure, 'W'))
#---translate the water box
gmx('editconf',
structure=structure,
gro='solvate-water-shifted',
flag='-translate 0 0 %f' %
(wordspace['bilayer_dimensions_slab'][2] / 2.),
log='editconf-solvate-shift')
#---combine and trim with new box vectors
#---! skipping minimization?
structure = 'solvate-water-shifted'
boxdims_old, boxdims = get_box_vectors(structure)
boxvecs = wordspace['bilayer_dimensions_slab'][:2] + [
wordspace['bilayer_dimensions_slab'][2] + boxdims[2]
]
gro_combinator('%s.gro' % incoming_structure,
structure,
box=boxvecs,
cwd=wordspace['step'],
gro='solvate-dense')
structure = 'solvate-dense'
trim_waters(structure=structure,
gro='solvate',
boxcut=False,
gap=wordspace['protein_water_gap'],
method='cgmd',
boxvecs=boxvecs)
structure = 'solvate'
nwaters = count_molecules(structure, 'W')
component('W', count=nwaters)
wordspace['bilayer_dimensions_solvate'] = boxvecs
wordspace['water_without_ions'] = nwaters