def __init__(self, input_file):
self.multiple_allowed = ['conformation', 'plugin']
# read the input file
with open(input_file, 'r', encoding=encoding) as f:
self.parameters = dict()
for line in f.readlines():
# the last check seems to eager.
# The line split as the iterable should catch that as a new line
if not any((line.startswith('#'), line.startswith(' '),
line.startswith('\n'))):
frags = []
for i in line.split():
if i.startswith("#"):
break
frags.append(i.strip())
if frags[
0] not in self.parameters: # that means if the key appears for the first time
if frags[
0] in self.multiple_allowed: # i.e. if the key is 'conformation'
self.parameters[frags[0]] = []
self.parameters[frags[0]].append(' '.join(
frags[1:]))
else:
self.parameters[frags[0]] = ' '.join(frags[1:])
else:
if frags[0] in self.multiple_allowed:
self.parameters[frags[0]].append(' '.join(
frags[1:]))
else:
self.write2output(
'Ignoring second definition of: {}'.format(
frags[0]))
# check which enviroment is used
if "enviroment" not in self.parameters:
self.write2output(
"DEFAULT: PACMAN assumes standard local configuration.\n")
self.write2output(
"DEFAULT: Modules are not loaded using the `module` command.\n"
)
self.write2output(
"DEFAULT: To use PACMAN using modules set `enviroment` to `cluster`.\n"
)
self.parameters['enviroment'] = 'local'
# essential self.parameters that have to be present
essential_conf = (
'pdbname',
'psfname',
'segname',
'parameter_path',
)
check_econf = [i in self.parameters for i in essential_conf]
if all(check_econf):
self.parameters['enzyme_resids_list'] = self.create_pdb_resid_list(
self.parameters['segname'])
self.parameters['prms'] = glob.glob(
os.path.join(self.parameters['parameter_path'], '*.prm'))
self.parameters['rtfs'] = glob.glob(
os.path.join(self.parameters['parameter_path'], '*.rtf'))
self.parameters['strs'] = glob.glob(
os.path.join(self.parameters['parameter_path'], '*.str'))
self.parameters['prms'] += self.parameters['strs']
self.parameters['rtfs'] += self.parameters['strs']
# if not all essential keywords were in the main input file, check for supplemental input files
else:
if "conformation" in self.parameters:
self.write2output(
('INFO: Reading parameters for conformation from {}\n' +
'').format(self.parameters['conformation']))
self.parameters["conf_par"] = list()
for conformation in self.parameters['conformation']:
self.parameters["conf_par"].append(
self.read_conformation(conformation))
# normalize weights for conformations so that the sum of absolute weights is 1
weights = [
confo['weight'] for confo in self.parameters["conf_par"]
]
abso = [abs(i) for i in weights]
norm = [i / sum(abso) for i in weights]
for i in range(len(self.parameters["conf_par"])):
self.parameters["conf_par"][i]['weight'] = norm[i]
self.parameters['enzyme_resids_list'] = self.parameters[
"conf_par"][0]['enzyme_resids_list']
# TODO: Think about this workaround
self.parameters['rtfs'] = self.parameters["conf_par"][0][
'rtfs']
else:
mes = "ERROR: The parameter(s) `{}` is/are missing or misspelled but crucial!"
mes = mes.format(", ".join(i for i in essential_conf
if i not in self.parameters))
raise KeyError(mes)
# set number of cores used by PACMAN
if 'cpus' in self.parameters:
self.parameters2int('cpus')
else:
self.parameters['cpus'] = 1
self.write2output(
'DEFAULT: NAMD simulations will be run on a single core!\n')
# set maximum number of attempted mutations
if "cycles" in self.parameters:
self.parameters2int('cycles')
else:
self.parameters['cycles'] = 10
self.write2output('DEFAULT: PACMAN will run for 10 cycles!\n')
# set maximum change in total charge of the system
if "max_charge_change" in self.parameters:
self.parameters2int('max_charge_change')
else:
self.parameters['max_charge_change'] = np.inf
self.write2output(
'DEFAULT: The change in charge during the design process is unlimited!\n'
)
# set maximum number of accepted mutations
if "max_mutations" in self.parameters:
self.parameters2int('max_mutations')
else:
self.parameters['max_mutations'] = 10
self.write2output(
'DEFAULT: A maximum of 10 mutations will be performed!\n')
# set name of main output file
if "logfile" not in self.parameters:
self.parameters['logfile'] = 'design.log'
self.write2output('DEFAULT: Output will be written to {}\n'.format(
self.parameters['logfile']))
# set force with which dihedrals are restrained to optimal values during placment
if "dihedral_force" in self.parameters:
self.parameters2int('dihedral_force')
else:
self.parameters['dihedral_force'] = 90
self.write2output('DEFAULT: A force constant of 90 will be' +
' used for dihedral restraints.')
# set temperature in the Boltzmann factor of the Metropolis-Monte-Carlo criterion
if 'MC_temperature' in self.parameters:
self.parameters2float('MC_temperature')
else:
self.parameters['MC_temperature'] = 500
self.write2output('DEFAULT: PACMAN will use a temperature of 500' +
' K for MC steps.')
# set probability of single mutations
if 'p_single' in self.parameters:
self.parameters2float('p_single')
else:
self.parameters['p_single'] = 1.0
self.write2output(
'DEFAULT: Using {} for single mutation probabilities.\n'.
format(self.parameters['p_single']))
p_total = self.parameters['p_single']
# set probability of double mutations
if 'p_double' in self.parameters:
self.parameters2float('p_double')
else:
self.parameters['p_double'] = 0.0
self.write2output(
'DEFAULT: Using {} for double mutation probabilities.\n'.
format(self.parameters['p_double']))
p_total += self.parameters['p_double']
# set probabilty of triple mutations
if 'p_triple' in self.parameters:
self.parameters2float('p_triple')
self.write2output(
'DEFAULT: Using {} for triple mutation probabilities.\n'.
format(self.parameters['p_triple']))
else:
self.parameters['p_triple'] = 0.0
p_total += self.parameters['p_triple']
# check whether probabilitis add up to 1, are positive deifinite and inidividually smaller than 1
if any([(i < 0 or i > 1)
for i in (p_total, self.parameters['p_single'],
self.parameters['p_double'],
self.parameters['p_triple'])]) \
or abs(1 - p_total) > 1e-6:
raise ValueError(
"ERROR: Please consider the axioms of Kolmogorov when" +
"choosing your mutation probabilities. Thank you.\n")
# read amino acids defined in rtfs
full = []
print(self.parameters['rtfs'])
for item in self.parameters['rtfs']:
print(item)
with open(item, 'r', encoding=encoding) as topf:
for line in topf.readlines():
if line.startswith('RESI'):
full.append(line.split()[1].strip())
print(full)
full = list(set(full))
# set set of amino acids which can be mutation targets
if 'aa_set' in self.parameters:
natural = [
'ALA', 'ARG', 'ASN', 'ASP', 'CYS', 'GLN', 'GLU', 'GLY', 'HSD',
'ILE', 'LEU', 'LYS', 'MET', 'PHE', 'PRO', 'SER', 'THR', 'TRP',
'TYR', 'VAL'
]
# set of amino acids without CYS and PRO
if self.parameters['aa_set'] == 'soluble':
aminos = [
'ALA', 'ARG', 'ASN', 'ASP', 'GLN', 'GLU', 'GLY', 'HSD',
'ILE', 'LEU', 'LYS', 'MET', 'PHE', 'SER', 'THR', 'TRP',
'TYR', 'VAL'
]
# set of apolar amino acids without CYS and PRO
elif self.parameters['aa_set'] == 'transmembrane':
aminos = [
'ALA', 'ILE', 'LEU', 'MET', 'PHE', 'TRP', 'TYR', 'VAL'
]
# set of all amino acids
elif self.parameters['aa_set'] == 'natural':
aminos = natural
# use this with extreme caution if at all!
elif self.parameters['aa_set'] == 'full':
sys.stderr.write(
"WARNING: You are allowing all RESI entries in your rtfs as mutation targets!\n"
+
"WARNING: This is probably a very bad idea as it easily leads to problems.\n"
+
"WARNING: Consider specifying a list of allowed residues instead!\n"
+
"WARNING: Make sure all of them can fit in a protein chain and add them to\n"
+
"WARNING: the dictionary of dihedral angles if necessary! Good Luck!\n"
)
aminos = full
self.write2output(
"INFO: The following residues are possible mutation targets: \n"
)
self.write2output(full)
else:
aminos = [
i.strip() for i in self.parameters['aa_set'].split(',')
]
if not all([i in full for i in aminos]):
raise ValueError(
"ERROR: You can only specify either 'full', 'soluble', 'transmembrane' or a "
+
"comma separated list of the following amino acids: {}"
.format(' '.join(full)))
self.parameters['aa_set'] = aminos
else:
self.parameters['aa_set'] = [
'ALA', 'ARG', 'ASN', 'ASP', 'GLN', 'GLU', 'GLY', 'HSD', 'ILE',
'LEU', 'LYS', 'MET', 'PHE', 'SER', 'THR', 'TRP', 'TYR', 'VAL'
]
self.write2output(
'DEFAULT: PACMAN assumes all 20 natural amnino acids are legit targets for mutation\n'
+ 'DEFAULT: Please consult the manual for other options.\n')
check_list = [i in full for i in self.parameters['aa_set']]
if not all(check_list):
inexistents = [
bi for bi, ci in zip(self.parameters['aa_set'], check_list)
if not ci
]
raise ValueError(
'ERROR: kindly consider checking whether your rtf files contain all possible'
+
'mutation targets. Hint: These residues {} are not contained.\n'
.format(inexistents))
# set whether global side chain relaxation is allowed after mutation
if 'fixed_sidechains' in self.parameters:
self.noyesbool('fixed_sidechains')
else:
self.parameters['fixed_sidechains'] = False
self.write2output(
'DEFAULT: It is assumed that you do not want to restrain global adaption of'
+ 'your protein.\n')
# set whether an initial minimization is performed before the evaluation of the input structure
if 'minimize' in self.parameters:
self.noyesbool('minimize')
else:
self.parameters['minimize'] = True
self.write2output(
"DEFAULT: Your starting structure will be minimized before evaluation\n"
)
# generate a list of enzyme resids that will not be mutated
# set residues which should not be mutated
if 'immutables' in self.parameters:
self.parameters['immutables'] = self.parameter2residlist(
'immutables')
self.parameters['immutables_vmd'] = (
'resid {} and segname {}' + '').format(
' '.join([str(i) for i in self.parameters['immutables']]),
self.parameters['segname'])
# remove immutables from the resid list
if 'conf_par' in self.parameters:
for snap in self.parameters['conf_par']:
snap['enzyme_resids_list'] = list(
set(snap['enzyme_resids_list']).difference(
self.parameters['immutables']))
else:
self.parameters['enzyme_resids_list'] = list(
set(self.parameters['enzyme_resids_list']).difference(
self.parameters['immutables']))
else:
self.parameters['immutables'] = []
self.parameters['immutables_vmd'] = ''
# set the solvent used during namd simulations (affects which templates are selected
if 'solvent' in self.parameters:
self.parameter_check_enum('solvent',
['vacuum', 'GBIS', 'water', 'membrane'])
else:
self.parameters['solvent'] = 'vacuum'
self.write2output(
"DEFAULT: All MD simulations during the design process will be performed in vacuo.\n"
)
# set the criterion for Monte-Carlo evalutaion of the mutations
if 'MC_criterion' in self.parameters:
self.parameter_check_enum('MC_criterion',
['vdw', 'elec', 'nonbonded', 'external'])
# enable Plugins for advanced properties to be evaluated
if self.parameters['MC_criterion'] == 'external':
self.plugin_pars = {}
if 'plugin' not in self.parameters:
raise ValueError(
"ERROR: If you want to use an external MC criterion you have "
+ "to name the plugin to use." +
"Try again or read the README")
tmp_plugin = []
try:
for i in self.parameters['plugin']:
"""plugins should be specified as plugin _label_:_pluginname_ """
try:
label, plug = i.split(':')
except ValueError:
raise ValueError(
"Plugins should be specified as _label_:_pluginname_!"
)
plugin_par_func = Plugin.Plugin.get_plugin_pars(plug)
plugin_pars = {}
for par in plugin_par_func.keys():
par_w_label = "{}:{}".format(label, par)
try:
plugin_pars[par] = plugin_par_func[par][1](
self.parameters[par_w_label])
except KeyError:
if plugin_par_func[par][0]:
mes = 'Plugin `{}` is missing essential argument `{}`!'.format(
plug, par_w_label)
raise Plugin.PluginArgError(mes)
else:
plugin_pars[par] = plugin_par_func[par][1](
None)
try:
key = '{}:w'.format(label)
self.parameters2float(key)
plugin_pars['w'] = self.parameters[key]
if plugin_pars['w'] < 0:
mes = "ERROR: Plugin weights must be > 0!"
raise ValueError(mes)
if plugin_pars['w'] == 0:
sys.stderr.write(
"WARNING: Ypu set a plugin weight to 0.0 . If you just want to"
+ " monitor a property" +
" without designing according to it,a weight of 0 might be useful,"
+
" IF the plugin is inexpensive! Do you really want to do this?\n"
)
except KeyError:
plugin_pars['w'] = None
tmp_plugin.append((plug, plugin_pars))
if any([t[1]['w'] is not None for t in tmp_plugin
]) and not all([t[1]['w'] for t in tmp_plugin]):
mes = "ERROR: Either all plugins have a specified weight attribute _label_:w or none."
raise RuntimeError(mes)
elif all([t[1]['w'] is None for t in tmp_plugin]):
for t in tmp_plugin:
t[1]['w'] = 1 / len(tmp_plugin)
else:
sum_w = sum([t[1]['w'] for t in tmp_plugin])
for t in tmp_plugin:
t[1]['w'] = t[1]['w'] / sum_w
except:
raise ImportError("plugin definition did not work...")
self.parameters['plugin'] = tmp_plugin
else:
self.parameters['plugin'] = []
else:
self.parameters['MC_criterion'] = 'vdw'
self.parameters['plugin'] = ' '
self.parameters['plugin_par'] = {}
self.write2output(
"DEFAULT: Mutations will be accepted or rejected" +
"according to the change in vdw-energy.\n")
# set the property-dependent threshold to be surpassed by the properties
if 'threshold' not in self.parameters:
if self.parameters['MC_criterion'] == 'vdw':
self.parameters['threshold'] = -2.5
elif self.parameters['MC_criterion'] == 'elec':
self.parameters['threshold'] = -20.0 * 4184
elif self.parameters['MC_criterion'] == 'nonbonded':
self.parameters['threshold'] = -50.0 * 4184
else: # only possibilty left is external
raise ValueError(
"ERROR: If you want to use an external MC criterion you have "
+
"to define a custom threshold. Try again or read the README"
)
else:
self.parameters['threshold'] = self.value2float(
self.parameters['threshold'])
# set inter-atomic distance restraints
if 'constraint_pairs' in self.parameters:
self.parameters['constraint_pairs'] = self.get_constraints_pairs()
self.parameters['constraint'] = True
if 'constraint_lengths' in self.parameters:
self.parameters[
'constraint_lengths'] = self.get_constraints_attr(
'constraint_lengths')
else:
self.parameters['constraint_lengths'] = [4] * len(
self.parameters['constraint_pairs'])
self.write2output(
'DEFAULT: PACMAN will use standard lengths of 0.4 nm for constraints.\n'
)
if 'constraint_forces' in self.parameters:
self.parameters[
'constraint_forces'] = self.get_constraints_attr(
'constraint_forces')
else:
self.parameters['constraint_forces'] = [200] * len(
self.parameters['constraint_pairs'])
self.write2output(
'DEFAULT: PACMAN will use a standard force constant of 200 for constraints.\n'
)
else:
self.parameters['constraint'] = False
self.parameters['constraint_pairs'] = []
self.parameters['constraint_forces'] = []
self.parameters['constraint_lengths'] = []
if 'prefix' not in self.parameters and 'conformation' not in self.parameters:
self.parameters[
'prefix'] = '' # self.parameters['pdbname'].split('.pdb')[0]
self.write2output(
'DEFAULT: Files will be named after the provided PDB file.\n')
self.write_full_config()
