Package schrodinger :: Package application :: Package matsci :: Package genetic_optimization :: Module genetic_optimization

[frames] | no frames]

Module genetic_optimization

Classes

[hide private]

Property
Manage a property to be used in a genetic optimization.

StructureGenome
Manage a genome.

Skip
Manage a skip.

Failure
Manage a failure.

CheckInput
Manage checking user input.

GeneticOptimization
Manage the genetic optimization.

Functions

[hide private]

set_title_to_stoichiometry(astructure, toappend=None, separation=STOICH_BASE_EXT_SEP)
Set the structure title to be the stoichiometry of the structure.

from_initial_population(genome, **args)
Draw a unique genome from the initial population.

int

get_num_simple_bonds(astructure)
Return the number of simple bonds in the provided structure.

schrodinger.structure.Structure

combine_two_structures(astructure, bstructure, offset=10.0)
Combine two structure objects into a single structure object using somewhat arbitrary placement.

tuple

bond_crossover(genome, **args)
Perform a crossover operation by swapping molecular fragments at two randomly choosen bonds, i.e.

dict

get_element_mutator_dict(astructure)
Return a dictionary where the keys contain the indicies of the mutatable atoms and the values contain those elements that the keyed atom may be mutated to.

list

get_isoelectronic_mutator_indicies(astructure)
Return a list of atom indicies that can be mutated by the isoelectronic mutator.

schrodinger.structure.Structure

get_child_like_parent(parent_st, children_sts, definition)
Return the child structure that is most like the provided parent.

int

elemental_mutator(genome, **args)
Perform a random elemental mutation to an element in the same column (as known as group) of the periodic table.

int

fragment_mutator(genome, **args)
Randomly mutate the genome by swapping a molecular fragement on one side of a bond by a similar fragment from a library.

int

isoelectronic_mutator(genome, **args)
Perform a random isoelectronic mutation from the following sets of series CH3X, NH2X, OHX, and FX, CH2XY, NHXY, OXY, and CHXYZ and NXYZ, where X, Y, and Z are non-H-bonds.

str

get_loggable_float(afloat, num_decimal=NUM_DECIMAL, field_width=FIELD_WIDTH)
Return a float as a string with the specified format.

uniquify_titles_callback(ga_obj)
Callback to uniquify titles of the individuals.

prepare_next_generation_dirs_callback(ga_obj)
Callback to update the generation property of the genomes and to create a subdirectory to hold the next series of evaluations.

manage_skips_callback(ga_obj)
Callback to manage skips in the evaluation.

manage_failures_callback(ga_obj)
Callback to manage failures in the evaluation.

logging_summary_callback(ga_obj)
Callback to log progress.

molecule_history_callback(ga_obj)
Callback to append all structures from all generations to individual log files.

bool

first_property(ga_obj)
Terminate when the first property has been matched.

bool

all_properties(ga_obj)
Terminate when all properties have been matched.

bool

unproductive(ga_obj)
Terminate if the maximum number of unproductive generations has been reached.

print_bad_jobs(all_bad_jobs, logger, bad_type='skip')
Log bad jobs, i.e.

str

get_output_file_name(basename)
Get the output file name from the basename.

str

get_generation_log_file_name(basename, generation)
Get the generation log file name.

function

thread_safe_evaluator(evaluator)
Decorator to make evaluator functions thread-safe, i.e.

float

get_structure_score(astructure, properties, conformational_search)
Return the structure score for the provided structure.

float

structure_evaluator(genome)
This is the structure evaulator.

float

base_evaluator(genome)
This is the base evaulator used to wrap all other evaluators.

launcher.Launcher

optoelectronics_evaluator(genome)
Run an optoelectronics job.

list

apply_uniform_operator_weights(operators)
Set the operator weights uniformly.

bool

structure_is_open_shell(astructure, ignore_charge=True)
Return True if the provided structure is open shell, i.e.

dict

get_element_histogram(astructure)
Return a dictionary where keys are elements and values are the numbers of atoms of a given element.

str

remove_basename_ext(stoich_ext)
Remove the basename extension from the given string and return the remainder which is the stoichiometry.

list of schrodinger.structure.Structure, int

get_low_energy_conformers(astructure_in, macromodel_options_file=None, remove_files=False, overwrite=False, seed=None)
Return the lowest energy conformers from a Macromodel conformational search.

schrodinger.structure.Structure or None

get_random_structure(structure_libs, tries_from_libs=TRIES_FROM_LIBS, structure_score_threshold=None, properties=None, conformational_search=CONFORMATIONAL_SEARCH)
From the given dictionary of libraries return a random structure.

schrodinger.structure.Structure or None

get_freezer_structure(structure_libs, tries_from_libs=TRIES_FROM_LIBS, structure_score_threshold=None, properties=None, conformational_search=CONFORMATIONAL_SEARCH, inoculate=NO_CHILD, crossover_applied=None, mutation_applied=None, basename_ext=None)
Return a random structure from the freezer and update that structure's properties.

Variables

[hide private]

__doc__ = ...

_version = '$Revision 0.0 $'

PYEVOLVE_LOG_EXT = '-pyevolve.log'

IN_MAE_EXT = '-in.mae'

OUT_MAE_EXT = '-out.mae'

BOND_CROSSOVER = 'bond'

CROSSOVER_CHOICES = [BOND_CROSSOVER]

DEFAULT_CROSSOVERS = [BOND_CROSSOVER]

ELEMENTAL_MUTATOR = 'elemental'

FRAGMENT_MUTATOR = 'fragment'

ISOELECTRONIC_MUTATOR = 'isoelectronic'

MUTATOR_CHOICES = [ELEMENTAL_MUTATOR, FRAGMENT_MUTATOR, ISOELE...

DEFAULT_MUTATORS = [FRAGMENT_MUTATOR, ISOELECTRONIC_MUTATOR]

GENERATIONS = 10

POPULATION = 8

CROSSOVER_RATE = 90.0

MUTATION_RATE = 90.0

RANK_SELECTION = 'rank'

ROULETTE_WHEEL_SELECTION = 'roulette_wheel'

TOURNAMENT_SELECTION = 'tournament'

TOURNAMENT_SELECTION_WITH_ROULETTE = 'tournament_with_roulette'

UNIFORM_SELECTION = 'uniform'

SELECTION_DICT = {RANK_SELECTION: Selectors.GRankSelector, ROU...

SELECTION_CHOICES = [RANK_SELECTION, ROULETTE_WHEEL_SELECTION,...

DEFAULT_SELECTION = 'roulette_wheel'

TOURNAMENT_SIZE = 2

UNPRODUCTIVE_TERM = 'unproductive'

FIRST_PROPERTY_TERM = 'first_property'

ALL_PROPERTIES_TERM = 'all_properties'

MAX_GENERATIONS_TERM = 'max_generations'

TERM_CHOICES = [UNPRODUCTIVE_TERM, FIRST_PROPERTY_TERM, ALL_PR...

DEFAULT_TERMS = [UNPRODUCTIVE_TERM, ALL_PROPERTIES_TERM]

NUM_UNPRODUCTIVE = 6

LINEAR_SCALING = 'linear'

POWER_LAW_SCALING = 'power_law'

EXPONENTIAL_SCALING = 'exponential'

SATURATED_SCALING = 'saturated'

SIGMA_TRUNCATION_SCALING = 'sigma_truncation'

BOLTZMANN_SCALING = 'boltzmann'

SCALING_DICT = {LINEAR_SCALING: Scaling.LinearScaling, POWER_L...

SCALING_CHOICES = [LINEAR_SCALING, POWER_LAW_SCALING, EXPONENT...

DEFAULT_SCALING = 'sigma_truncation'

ALLOWS_NEGATIVE_SCORES = [EXPONENTIAL_SCALING, SATURATED_SCALI...

ELITISM = 1

RANDOM_SEED = None
hash(x)

RANDOM_INT_BOUND = 1000000

NO_MINIMIZE = False
hash(x)

INDIVIDUAL_KEY = 'i_matsci_individual_index'

GENERATION_KEY = 'i_matsci_generation'

STRUCTURE_SCORE_KEY = 'r_matsci_structure_score'

RAW_SCORE_KEY = 'r_matsci_raw_score'

FIT_SCORE_KEY = 'r_matsci_fit_score'

SKIP_KEY = 'b_matsci_skipped'

FAILURE_KEY = 'b_matsci_failed'

LOCALHOST = 'localhost'

TPP_GA = 1

TPP_EVAL = 1

TPP_STR = '-TPP'

DEFAULT_EVAL_KWARGS = {}

ORGANIC = 'organic'

N_HETEROCYCLES = 'N-heterocycles'

O_HETEROCYCLES = 'O-heterocycles'

S_HETEROCYCLES = 'S-heterocycles'

MIXED_HETEROCYCLES = 'Mixed-heterocycles'

COMBIGLIDE_DEFAULT = 'combiglide_default'

OPTOELECTRONICS = 'optoelectronics'

ALL = 'all'

MMSHARE_MAIN_DATA = hunt('mmshare', dir= 'data')

FRAGMENT_LIBS = {ORGANIC: os.path.join(MMSHARE_MAIN_DATA, 'res...

FRAGMENT_LIBS_DEFAULT = [OPTOELECTRONICS]

ENTRY_NAME_KEY = 's_m_entry_name'

GROW_NAME_KEY = 's_m_grow_name'

PDB_ATOM_NAME_KEY = 's_m_pdb_atom_name'

PDB_RES_NAME_KEY = 's_m_pdb_residue_name'

CROSSOVER_PARENTS_KEY = 's_matsci_crossover_parents'

CROSSOVER_APPLIED_KEY = 's_matsci_crossover_applied'

MUTATION_PARENT_KEY = 's_matsci_mutation_parent'

MUTATION_APPLIED_KEY = 's_matsci_mutation_applied'

EVALUATOR_JOBS_DIR = 'evaluator_jobs'

GENER_SUBDIR = 'generation_'

NUM_DECIMAL = '%.2f'

FIELD_WIDTH = 10

INFINITE_SCORE = 1000000000

BAD_SCORE = -10000000

RXN_KEY = 's_matsci_RXN_representation'

SMALL_CHILD_FREQ = 0.75

SIMILAR_STDEV_CHILDREN_FREQ = 0.75

FILE_BASE_NAME = 'genopt'

NO_OPEN_SHELL = False
hash(x)

TERM_THRESH = 0.0001

SMARTS_PATTERN_SEPARATOR = '_'

SMARTS_PROP = 'smarts'

MOL_WEIGHT_PROP = 'molecular_weight'

NATOMS_PROP = 'natoms'

NELEMENTS_PROP = 'nelements'

STRUCTURE_PROPERTIES = {SMARTS_PROP, MOL_WEIGHT_PROP, NATOMS_P...

SMARTS_KEY = 'i_matsci_SMARTS_property_%s'

MOL_WEIGHT_KEY = 'r_m_Molecular_weight'

NATOMS_KEY = 'i_m_Number_of_atoms'

NELEMENTS_KEY = 'i_m_Number_of_elements'

NONE = 'none'

ATYPICAL = 'atypical'

ATYPICAL_PATTERNS = OrderedDict([('three-atom or longer chains...

ATYP_MW_TARGET = 1100

ATYP_MW_WEIGHT = 1.0

ATYP_NAT_TARGET = 200

ATYP_NAT_WEIGHT = 1.0

ATYP_NEL_TARGET = 5

ATYP_NEL_WEIGHT = 15

ATYPICAL_PROPS = ['name=natoms target=%d comparator=lt weight=...

DIHYDROGEN = 'dihydrogen'

DIHYDROGEN_PATTERNS = OrderedDict([('dihydrogen molecule', '[H...

DIHYDROGEN_PROPS = ['name=smarts patterns=%s minimax=min weigh...

PENALIZE_PROTOCOLS = OrderedDict([(ATYPICAL, (ATYPICAL_PATTERN...

PENALIZE_CHOICES = [NONE, ATYPICAL, DIHYDROGEN]

PENALIZE_DEFAULT = [ATYPICAL, DIHYDROGEN]

STRUCTURE_SCORE_THRESHOLD = -50.0

EXTS_TO_RETURN = ['.spm']

STOICH_BASE_EXT_SEP = '.'

CONFORMATIONAL_SEARCH = False
hash(x)

CONF_SEARCH_FAILED_KEY = 'b_matsci_Conf_Search_Failed'

CONF_SEARCH_SEED_RANGE = [1, 78593]

GENER_LOG_FILE_SEP = '-'

REMAINDER = 'remainder'

PREVIOUS = 'previous'

FRAGMENT = 'fragment'

FREEZER_CHOICES = [REMAINDER, PREVIOUS, FRAGMENT]

FREEZERS_DEFAULT = [REMAINDER, PREVIOUS]

FROM_FREEZER_KEY = 's_matsci_From_Freezer'

REMAINDER_FILE_EXT = '-remainder.mae'

NO_CHILD = 'no_child'

BAD_STRUCTURE = 'bad_structure'

INOCULATE_CHOICES = [NONE, NO_CHILD, BAD_STRUCTURE]

INOCULATE_DEFAULT = [NO_CHILD, BAD_STRUCTURE]

INOCULATE_KEY = 's_matsci_Inoculate'

TRIES_FROM_LIBS = 3

TRIES_FRAGMENT_MUTATOR = 3

Function Details

[hide private]

set_title_to_stoichiometry(astructure, toappend=None, separation=STOICH_BASE_EXT_SEP)

Set the structure title to be the stoichiometry of the structure.

Parameters:

astructure (schrodinger.structure.Structure) - the structure
toappend (str) - a string to append to the stoichiometry
separation (str) - used to separate the stoichiometry and the toappend str

from_initial_population(genome, **args)

Draw a unique genome from the initial population.

Parameters:

genome (StructureGenome) - a genome
args (dict) - dictionary of genetic optimization parameters created and used by pyevolve

get_num_simple_bonds(astructure)

Return the number of simple bonds in the provided structure. The definition of a simple bond follows from that used in the reaction channel module and is an acyclic single order bond that may involve a hydrogen atom.

Parameters:

astructure (schrodinger.structure.Structure) - the structure for which to get the number of simple bonds

Returns: int

the number of simple bonds

combine_two_structures(astructure, bstructure, offset=10.0)

Combine two structure objects into a single structure object using somewhat arbitrary placement.

Parameters:

astructure (schrodinger.structure.Structure) - the first of the structures to be combined
bstructure (schrodinger.structure.Structure) - the second of the structures to be combined
offset (float) - the final distance between the structures will be the sum of the molecular VDW radii plus this offset in Angstrom

Returns: schrodinger.structure.Structure

the combined structure object

bond_crossover(genome, **args)

Perform a crossover operation by swapping molecular fragments at two randomly choosen bonds, i.e. a double displacement reaction channel.

Parameters:

genome (StructureGenome) - a genome
args (dict) - dictionary of genetic optimization parameters created and used by pyevolve

Returns: tuple

tuple containing the sister and brother StructureGenome

get_element_mutator_dict(astructure)

Return a dictionary where the keys contain the indicies of the mutatable atoms and the values contain those elements that the keyed atom may be mutated to.

Parameters:

astructure (schrodinger.structure.Structure) - the structure to be mutated

Returns: dict

keys are atom indicies of those atoms that are mutatable and values are those elements that the atom can be mutated to

get_isoelectronic_mutator_indicies(astructure)

Return a list of atom indicies that can be mutated by the isoelectronic mutator.

Parameters:

astructure (schrodinger.structure.Structure) - the structure to be mutated

Returns: list

mutatable indicies

get_child_like_parent(parent_st, children_sts, definition)

Return the child structure that is most like the provided parent.

Parameters:

parent_st (schrodinger.structure.Structure) - the parent structure
children_sts (list of schrodinger.structure.Structure) - the children structures
definition (two-element list) - each sublist contains two atom indicies describing the reactive bonds in parent and fragment structures which created the children

Returns: schrodinger.structure.Structure

the sought child structure

elemental_mutator(genome, **args)

Perform a random elemental mutation to an element in the same column (as known as group) of the periodic table. Note that hydrogen and the halogens are considered to belong to the same column.

Parameters:

genome (StructureGenome) - a genome
args (dict) - dictionary of genetic optimization parameters created and used by pyevolve

Returns: int

the number of mutations applied, appears to never be used in PyEvolve

fragment_mutator(genome, **args)

Randomly mutate the genome by swapping a molecular fragement on one side of a bond by a similar fragment from a library.

Parameters:

genome (StructureGenome) - a genome
args (dict) - dictionary of genetic optimization parameters created and used by pyevolve

Returns: int

the number of mutations applied, appears to never be used in PyEvolve

isoelectronic_mutator(genome, **args)

Perform a random isoelectronic mutation from the following sets of series CH3X, NH2X, OHX, and FX, CH2XY, NHXY, OXY, and CHXYZ and NXYZ, where X, Y, and Z are non-H-bonds.

Parameters:

genome (StructureGenome) - a genome
args (dict) - dictionary of genetic optimization parameters created and used by pyevolve

Returns: int

the number of mutations applied, appears to never be used in PyEvolve

get_loggable_float(afloat, num_decimal=NUM_DECIMAL, field_width=FIELD_WIDTH)

Return a float as a string with the specified format.

Parameters:

afloat (float) - a float to convert to a string
num_decimal (str) - the format of the string representation
field_width (int) - the field width of the final string

Returns: str

the float as a string

uniquify_titles_callback(ga_obj)

Callback to uniquify titles of the individuals.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

prepare_next_generation_dirs_callback(ga_obj)

Callback to update the generation property of the genomes and to create a subdirectory to hold the next series of evaluations.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

manage_skips_callback(ga_obj)

Callback to manage skips in the evaluation.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

manage_failures_callback(ga_obj)

Callback to manage failures in the evaluation.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

logging_summary_callback(ga_obj)

Callback to log progress.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

molecule_history_callback(ga_obj)

Callback to append all structures from all generations to individual log files.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

first_property(ga_obj)

Terminate when the first property has been matched.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

Returns: bool

True to terminate, False otherwise

all_properties(ga_obj)

Terminate when all properties have been matched.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

Returns: bool

True to terminate, False otherwise

unproductive(ga_obj)

Terminate if the maximum number of unproductive generations has been reached.

Parameters:

ga_obj (GSimpleGA.GSimpleGA) - the entire current state of the genetic optimization

Returns: bool

True to terminate, False otherwise

print_bad_jobs(all_bad_jobs, logger, bad_type='skip')

Log bad jobs, i.e. skips and failures.

Parameters:

all_bad_jobs (dict) - a collection of bad subjobs, keys are genetic optimization generation and values are a list of Skip or Failure objects for bad subjobs
logger (logging.Logger) - output logger
bad_type (str) - specifies either 'skip' or 'fail' type

get_output_file_name(basename)

Get the output file name from the basename.

Parameters:

basename (str) - base name to use

Returns: str

output_file_name, name of output file

get_generation_log_file_name(basename, generation)

Get the generation log file name.

Parameters:

basename (str) - base name to use
generation (int) - the generation

Returns: str

generation_log_file_name, name of generation log file

thread_safe_evaluator(evaluator)

Decorator to make evaluator functions thread-safe, i.e. to make using the random module safe, meaning usability and reproducibility, with the multiprocessing module.

Parameters:

evaluator (function) - the function to decorate

Returns: function

the decorated function

get_structure_score(astructure, properties, conformational_search)

Return the structure score for the provided structure.

Parameters:

astructure (schrodinger.structure.Structure) - the structure to score
properties (list of Property) - the properties used in scoring
conformational_search (bool or str) - specifies whether a Macromodel conformational search will be performed prior to evaluation, when a string it specifies a simplified Macromodel input file containing extra options

Returns: float

the structure score

structure_evaluator(genome)

This is the structure evaulator.

Parameters:

genome (StructureGenome) - a genome

Returns: float

the score for this individual

Decorators:

@thread_safe_evaluator

base_evaluator(genome)

This is the base evaulator used to wrap all other evaluators.

Parameters:

genome (StructureGenome) - a genome

Returns: float

the score for this individual

Decorators:

@thread_safe_evaluator

optoelectronics_evaluator(genome)

Run an optoelectronics job.

Parameters:

genome (StructureGenome) - a genome

Returns: launcher.Launcher

the script launcher object for this individual, it is run in the base evaluator

apply_uniform_operator_weights(operators)

Set the operator weights uniformly.

Parameters:

operators (list) - a list of two-element tuples, each tuple contains first an operator function and second a weight

Returns: list

list of two-element tuples of operators and uniform weights

structure_is_open_shell(astructure, ignore_charge=True)

Return True if the provided structure is open shell, i.e. has an odd number of electrons.

Parameters:

astructure (schrodinger.structure.Structure) - the structure in question
ignore_charge (bool) - if True then ignore any structure.formal_charge settings

Returns: bool

True if the provided structure is open shell, False otherwise

get_element_histogram(astructure)

Return a dictionary where keys are elements and values are the numbers of atoms of a given element.

Parameters:

astructure (schrodinger.structure.Structure) - the structure in question

Returns: dict

dictionary with element histogram, keys are elements (strs) and values are numbers (ints)

remove_basename_ext(stoich_ext)

Remove the basename extension from the given string and return the remainder which is the stoichiometry. Do this instead of having to recompute the stoichiometry which can be expensive.

Parameters:

stoich_ext (str) - contains the stoichiometry and basename extension

Returns: str

stoichiometry

get_low_energy_conformers(astructure_in, macromodel_options_file=None, remove_files=False, overwrite=False, seed=None)

Return the lowest energy conformers from a Macromodel conformational search.

Parameters:

astructure_in (schrodinger.structure.Structure) - the structure to search for conformations
macromodel_options_file (str or None) - the name of a simplified Macromodel input file that contains any options to use in addition to those used by default in a conformational search or None if there are none and you just want to use the defaults
remove_files (bool) - if the job is successful, specifies whether to remove all files created for it after it finishes
overwrite (bool) - if True then the coordinates of the input structure will be overwritten by those of the lowest energy conformer and that structure alone returned by this function
seed (int or None) - used to seed the random number generator used in the Macromodel conformational search, should be in CONF_SEARCH_SEED_RANGE, if None then if a CONFSEARCH_SEED has been specified in macromodel_options_file it will be used, otherwise the system time in seconds will be used

Returns: list of schrodinger.structure.Structure, int

the structures of the lowest energy conformers sorted by increasing energy and the seed used in the conformational search (same as input if input was given either as seed or in macromodel_options_file)

get_random_structure(structure_libs, tries_from_libs=TRIES_FROM_LIBS, structure_score_threshold=None, properties=None, conformational_search=CONFORMATIONAL_SEARCH)

From the given dictionary of libraries return a random structure.

Parameters:

structure_libs (dict) - keys are strings specifying the types of libraries to be used and can be module constants from FREEZER_CHOICES.keys(), values are lists of libraries by type and can be either module constants from FRAGMENT_LIBS.keys(), ALL, or the names of Maestro files (including the file extensions)
tries_from_libs (int) - the number of times to try before giving up
structure_score_threshold (float or None) - specifies that a structure with a structure score greater-than-or-equal-to this threshold is sought, the best of the considered structures will be returned and will contain several structure properties related to the scoring
properties (list of Property or None) - the properties used in structure scoring
conformational_search (bool or str) - specifies whether a Macromodel conformational search will be performed prior to evaluation, when a string it specifies a simplified Macromodel input file containing extra options

Returns: schrodinger.structure.Structure or None

the random structure or None if one couldn't be found

get_freezer_structure(structure_libs, tries_from_libs=TRIES_FROM_LIBS, structure_score_threshold=None, properties=None, conformational_search=CONFORMATIONAL_SEARCH, inoculate=NO_CHILD, crossover_applied=None, mutation_applied=None, basename_ext=None)

Return a random structure from the freezer and update that structure's properties.

Parameters:

structure_libs (dict) - keys are strings specifying the types of libraries to be used and can be module constants from FREEZER_CHOICES.keys(), values are lists of libraries by type and can be either module constants from FRAGMENT_LIBS.keys(), ALL, or the names of Maestro files (including the file extensions)
tries_from_libs (int) - the number of times to try before giving up
structure_score_threshold (float or None) - specifies that a structure with a structure score greater-than-or-equal-to this threshold is sought, the best of the considered structures will be returned and will contain several structure properties related to the scoring
properties (list of Property or None) - the properties used in structure scoring
conformational_search (bool or str) - specifies whether a Macromodel conformational search will be performed prior to evaluation, when a string it specifies a simplified Macromodel input file containing extra options
inoculate (str) - specify the reason for drawing from the freezer, which is an inoculate option from INOCULATE_CHOICES
crossover_applied (str or None) - specify the intended crossover operator or None if there isn't to be one
mutation_applied (str or None) - specify the intended mutation operator or None if there isn't to be one
basename_ext (str or None) - specify an extension to append to the stoichiometry which is used to set the title of the returned structure

Returns: schrodinger.structure.Structure or None

the random structure or None if one couldn't be found

Variables Details

[hide private]

doc

Value:

"""
Classes and functions for the genetic optimization module.

Copyright Schrodinger, LLC.  All rights reserved."""

MUTATOR_CHOICES

Value:

[ELEMENTAL_MUTATOR, FRAGMENT_MUTATOR, ISOELECTRONIC_MUTATOR]

SELECTION_DICT

Value:

{RANK_SELECTION: Selectors.GRankSelector, ROULETTE_WHEEL_SELECTION: Se
lectors.GRouletteWheel, TOURNAMENT_SELECTION: Selectors.GTournamentSel
ectorAlternative, TOURNAMENT_SELECTION_WITH_ROULETTE: Selectors.GTourn
amentSelector, UNIFORM_SELECTION: Selectors.GUniformSelector}

SELECTION_CHOICES

Value:

[RANK_SELECTION, ROULETTE_WHEEL_SELECTION, TOURNAMENT_SELECTION, TOURN
AMENT_SELECTION_WITH_ROULETTE, UNIFORM_SELECTION]

TERM_CHOICES

Value:

[UNPRODUCTIVE_TERM, FIRST_PROPERTY_TERM, ALL_PROPERTIES_TERM, MAX_GENE
RATIONS_TERM]

SCALING_DICT

Value:

{LINEAR_SCALING: Scaling.LinearScaling, POWER_LAW_SCALING: Scaling.Pow
erLawScaling, EXPONENTIAL_SCALING: Scaling.ExponentialScaling, SATURAT
ED_SCALING: Scaling.SaturatedScaling, SIGMA_TRUNCATION_SCALING: Scalin
g.SigmaTruncScaling, BOLTZMANN_SCALING: Scaling.BoltzmannScaling}

SCALING_CHOICES

Value:

[LINEAR_SCALING, POWER_LAW_SCALING, EXPONENTIAL_SCALING, SATURATED_SCA
LING, SIGMA_TRUNCATION_SCALING, BOLTZMANN_SCALING]

ALLOWS_NEGATIVE_SCORES

Value:

[EXPONENTIAL_SCALING, SATURATED_SCALING, SIGMA_TRUNCATION_SCALING, BOL
TZMANN_SCALING]

FRAGMENT_LIBS

Value:

{ORGANIC: os.path.join(MMSHARE_MAIN_DATA, 'res', 'organic.bld'), N_HET
EROCYCLES: os.path.join(MMSHARE_MAIN_DATA, 'res', 'N-heterocycles.bld'
), O_HETEROCYCLES: os.path.join(MMSHARE_MAIN_DATA, 'res', 'O-heterocyc
les.bld'), S_HETEROCYCLES: os.path.join(MMSHARE_MAIN_DATA, 'res', 'S-h
eterocycles.bld'), MIXED_HETEROCYCLES: os.path.join(MMSHARE_MAIN_DATA,
 'res', 'Mixed-heterocycles.bld'), COMBIGLIDE_DEFAULT: os.path.join(MM
SHARE_MAIN_DATA, 'cg', 'interactive_palettes', 'default_palette.mae'),
 OPTOELECTRONICS: os.path.join(MMSHARE_MAIN_DATA, 'genetic_optimizatio
...

STRUCTURE_PROPERTIES

Value:

{SMARTS_PROP, MOL_WEIGHT_PROP, NATOMS_PROP, NELEMENTS_PROP}

ATYPICAL_PATTERNS

Value:

OrderedDict([('three-atom or longer chains containing N, O, P, and S',
 '[$([#7,#8,#15,#16][#7,#8,#15,#16][#7,#8,#15,#16])]'), ('two or more 
different halogens', 'F.Cl_F.Br_F.I_Cl.Br_Cl.I_Br.I')])

ATYPICAL_PROPS

Value:

['name=natoms target=%d comparator=lt weight=%.1f' %(ATYP_NAT_TARGET, 
ATYP_NAT_WEIGHT), 'name=molecular_weight target=%d comparator=lt weigh
t=%.1f' %(ATYP_MW_TARGET, ATYP_MW_WEIGHT), 'name=nelements target=%d c
omparator=lt weight=%.1f' %(ATYP_NEL_TARGET, ATYP_NEL_WEIGHT), 'name=s
marts patterns=%s minimax=min weight=15.0' % '_'.join(ATYPICAL_PATTERN
S.values())]

DIHYDROGEN_PATTERNS

Value:

OrderedDict([('dihydrogen molecule', '[H][H]')])

DIHYDROGEN_PROPS

Value:

['name=smarts patterns=%s minimax=min weight=100.0' % '_'.join(DIHYDRO
GEN_PATTERNS.values())]

PENALIZE_PROTOCOLS

Value:

OrderedDict([(ATYPICAL, (ATYPICAL_PATTERNS, ATYPICAL_PROPS)), (DIHYDRO
GEN, (DIHYDROGEN_PATTERNS, DIHYDROGEN_PROPS))])