load_modify_sample_utils
Functions
|
Function to load and return a model in cobra format from given path. |
|
Function to find and return the objective functions of a cobra model |
|
Function to create and return a dictionary of a cobra's model reaction bounds. |
|
Function that modifies a given cobra model. |
|
Function that performs sampling with the OptGP sampler of cobrapy. Users can define number of samples, |
|
Function that performs sampling with dingo. Users can define the ess and psrf parameters. |
|
Function that performs sampling with the gapsplit algorithm. Users can define number of samples, if they want |
|
Function that exports samples to a pickle file |
|
Function that loads samples from a pickle file |
|
Function that plots a grid of the sampling distributions of all model reactions. |
|
Function that prints statistics for the sampling distribution of a specified model reaction |
Module Contents
- load_modify_sample_utils.load_model(filepath: str) Tuple[cobra.Model, List[cobra.Reaction], List[str]]
Function to load and return a model in cobra format from given path. It returns 3 objects, the cobra model and 2 lists, one with the model reactions (cobra_reactions) and one with their IDs (cobra_reactions_ids)
Keyword arguments: filepath – path to cobra model in sbml format
Returns: Tuple[Model, List, List]:
cobra_model – metabolic model in cobra format cobra_reactions – list with the model reactions cobra_reactions_ids – list with the string IDs of model reactions
- load_modify_sample_utils.get_objective_functions(cobra_model: cobra.Model) List[str]
Function to find and return the objective functions of a cobra model
Keyword arguments: cobra_model (Model) – cobra model object
Returns: objective_functions_ids (List) – List containing one or multiple objective functions of given model
- load_modify_sample_utils.get_reaction_bounds(cobra_model: cobra.Model) Dict[str, Tuple[float, float]]
Function to create and return a dictionary of a cobra’s model reaction bounds. Reaction IDs are keys and their corresponding bounds are values.
Keyword arguments: cobra_model (Model) – cobra model object
Returns: reaction_bounds_dict (Dict) – Dictionary containing reaction IDs as keys and their bounds as values
- load_modify_sample_utils.modify_model(cobra_model: cobra.Model, objective_function: str = None, optimal_percentage: float = 100, reaction_bounds: dict = {}, objective_direction: str = 'max') cobra.Model
Function that modifies a given cobra model.
Users can change objective function, optimal percentage, define custom reaction bounds and change objective direction. Lower bound of the objective will be fixed based on the optimal percentage, in all cases.
Keyword arguments: cobra_model (Model) – cobra model object objective_function (str) – string with the requested objective function ID to be assigned to the model optimal_percentage (float) – percentage of the FBA solution to be set as lower bound to the objectie function. reaction_bounds (dict) – dictionary of custom reaction bounds to be assigned to the model’s reactions objective_direction (str) – the direction to optimize the model: max and min are available options.
Returns: modified_cobra_model (Model) – A modified cobra model object
- load_modify_sample_utils.sample_optgp(cobra_model: cobra.Model, n_samples: int = 3000, thinning: int = 100, reaction_in_rows: bool = True) numpy.typing.NDArray[numpy.float64]
Function that performs sampling with the OptGP sampler of cobrapy. Users can define number of samples, thinning parameter. Resulting samples dataset is returned in a numpy array with reactions as rows or columns specified by user
Keyword arguments: cobra_model (Model) – cobra model object n_samples (int) – defines the number of samples returned to the user. thinning (int) – defines the thinning parameter. Default to 100 means samples are returned every 100th step. reaction_in_rows (bool) – boolean variable that if True reactions are rows in the sampling dataframe.
Returns: samples_optgp (NDArray[np.float64]) – Numpy 2D array of the sampling results
- load_modify_sample_utils.sample_dingo(dingo_model, ess: int = 1000, psrf: bool = False, solver: str = 'gurobi', final_n_samples: int = None, reaction_in_rows: bool = True) numpy.typing.NDArray[numpy.float64]
Function that performs sampling with dingo. Users can define the ess and psrf parameters. Resulting samples dataset is returned in a numpy array with reactions as rows or columns specified by user. User can call the “reduce_samples_dimensions” function to reduce the samples to a specified number
Keyword arguments: dingo_model – dingo model object ess – stands for the effective sample size (ESS) (default value is 1000). psrf – is a flag to request an upper bound equal to 1.1 for the value of the potential scale reduction factor of each marginal flux (default option is False). solver – defines the linear programming solver final_n_samples – integer that defines the final number of samples returned to the user (post sampling step) reaction_in_rows – boolean variable that if True reactions are rows in the sampling dataframe.
Returns: samples_dingo (NDArray[np.float64]) – Numpy 2D array of the sampling results
- load_modify_sample_utils.sample_gapsplit(cobra_model: cobra.Model, n_samples: int = 3000, reaction_in_rows: bool = True, add_loopless_cobrapy: bool = False, fraction_of_optimum: float = 0) numpy.typing.NDArray[numpy.float64]
Function that performs sampling with the gapsplit algorithm. Users can define number of samples, if they want to modify their model with the “add_loopless” function and a value for the “fraction_of_optimum” parameter used in the gapsplit algorithm. Resulting samples dataset is returned in a numpy array with reactions as rows or columns specified by user
Keyword arguments: cobra_model (Model) – cobra model object n_samples (int) – defines the number of samples returned to the user. reaction_in_rows (bool) – boolean variable that if True reactions are rows in the sampling dataframe add_loopless_cobrapy (bool) – Boolean variable that if True modifies the model with the add_loopless function fraction_of_optimum (float) – Float variable that defines the fraction_of_optimum parameter used in gapsplit algorithm
Returns: gapsplit_samples (NDArray[np.float64]) – Numpy 2D array of the sampling results
- load_modify_sample_utils.export_to_pickle(samples: numpy.typing.NDArray[numpy.float64], filename: str)
Function that exports samples to a pickle file
Keyword arguments: samples (NDArray[np.float64]) – Numpy 2D array of the samples filename (str) – String to name the created file
- load_modify_sample_utils.load_from_pickle(filename: str) numpy.typing.NDArray[numpy.float64]
Function that loads samples from a pickle file
Keyword arguments: filename (str) – String of the filename containing the samples
- load_modify_sample_utils.plot_grid_reactions_flux(samples: numpy.typing.NDArray[numpy.float64], model_reactions: List[str], tolerance: float = 0.001, nrows: int = 20, ncols: int = 5, title='Sampling Distributions')
Function that plots a grid of the sampling distributions of all model reactions. User can define distributions plotted per rows and columns
Keyword arguments: samples (NDArray[np.float64]) – Numpy 2D array of the samples model_reactions (List[str]) – A list containing strings of reaction IDs tolerance (float) – tolerance level to make distribution plot transparent based on flux range nrows (int) – Number of rows for the plot ncols (int) – Number of columns for the plot title (str) – Title of the plot
- load_modify_sample_utils.sampling_statistics(samples: numpy.typing.NDArray[numpy.float64], model_reactions: List[str] = None, reaction_id: str = '') Tuple[float, float, float, float, float, float]
Function that prints statistics for the sampling distribution of a specified model reaction
Keyword arguments: samples (NDArray[np.float64]) – Numpy 2D array of the samples model_reactions (List[str]) – A list containing strings of reaction IDs reaction_id (str) – Reaction ID to calculate sampling statistics on
Returns: Tuple[float, float, float, float, float, float]
mean (float) – average value from the reaction’s flux distribution min (float) – minimum value from the reaction’s flux distribution max (float) – maximum value from the reaction’s flux distribution std (float) – standard deviation value from the reaction’s flux distribution skewness (float) – skewness value from the reaction’s flux distribution kurtosis (float) – kurtosis value from the reaction’s flux distribution