distributions_comparison_utils

Functions

`significantly_altered_reactions`(], cobra_model, ...)	Function that takes as input at least 2 flux sampling conditions to compare and identifies significantly altered reactions
`plot_volcano`(pval_dict, fold_change_dict[, ...])	Creates a volcano plot to show results from differential flux analysis.
`dictionary_reaction_to_all_pathways`(→ Dict)	Function that builds a dictionary mapping each reaction to a list of all the pathway names it belongs.
`hypergeometric_test_pathway_enrichment`(→ pandas.DataFrame)	Function that perform a hypergeometric test to find significantly affected pathways between our sampling conditions.
`plot_pathway_enrichment`(hypergeometric_enrichment_df)	Function that takes as input the hypergeometric_pathway_enrichment_df dataframe created with the

Module Contents

distributions_comparison_utils.significantly_altered_reactions(conditions: List[numpy.typing.NDArray[numpy.float64]] = [], selected_comparisons: Tuple = [(0, 1)], cobra_model: cobra.Model = None, p_value_cutoff: float = 0.05, fold_change_cutoff: float = 1.2, std_cutoff: float = 0.01) → Tuple[List[str], List[str]]

Function that takes as input at least 2 flux sampling conditions to compare and identifies significantly altered reactions It performs a Kolmogorov-Smirnov (KS) non-parametric test and corrects p-value for multiple comparisons. It additionally calculates a fold change that together with the p-value classifies reactions as significantly altered or not.

Keyword arguments: conditions (List) – List of different flux sampling arrays selected_comparisons (List) – List showing which conditions to compare (useful when comparing more than 2 sampling arrays) cobra_model (Model) – cobra model object p_value_cutoff (float) – cutoff for p value from KS test to consider 2 distributions significantly different fold_change_cutoff (float) – cutoff for fold-change to consider 2 distributions significantly different std_cutoff (float) – cutoff to ensure distributions that are compared are not fixed to a certain value

Returns: Tuple[List, List]

significant_diff_reactions (List) – List containing reactions significantly altered between the conditions not_significant_diff_reactions (List) – List containing reactions not significantly altered between the conditions pval_dict (Dict) – Dictionary mapping reaction ID to corrected p-value. fold_change_dict (Dict) – Dictionary mapping reaction ID to fold-change value.

distributions_comparison_utils.plot_volcano(pval_dict: dict, fold_change_dict: dict, p_value_cutoff: float = 0.05, fold_change_cutoff: float = 1.2, annotate: list = [], width: int = 800, height: int = 600, title: str = '', show_cutoff_lines: bool = True)

Creates a volcano plot to show results from differential flux analysis.

Keyword arguments:: pval_dict (dict) – Dictionary mapping reaction ID to corrected p-value. fold_change_dict (dict) – Dictionary mapping reaction ID to fold-change value. p_value_cutoff (float) – Significance threshold for p-value. fold_change_cutoff (float) – Threshold for fold-change. annotate (list) – List of reaction IDs to annotate on the plot. width (int) – Width of the figure in pixels. height (int) – Height of the figure in pixels. title (str) – Title of the plot. show_cutoff_lines (bool) – Whether to draw p-value and fold-change cutoff lines.

distributions_comparison_utils.dictionary_reaction_to_all_pathways(df_kegg_pathways: pandas.DataFrame, reaction_col: str = 'model_reaction', pathway_col: str = 'pathway_names') → Dict

Function that builds a dictionary mapping each reaction to a list of all the pathway names it belongs.

Keyword arguments: df_kegg_pathways (pd.DataFrame): pandas DataFrame with reactions and pathway names. reaction_col: column name for reaction identifiers. pathway_col: column name for list of pathways.

Returns: reaction_to_pathways (dict): Dictionary with reaction IDs as keys and pathway names as values: {reaction_id: [pathway1, pathway2, …]}

distributions_comparison_utils.hypergeometric_test_pathway_enrichment(significant_reactions: List, model_reactions: List, reaction_to_pathways: Dict) → pandas.DataFrame

Function that perform a hypergeometric test to find significantly affected pathways between our sampling conditions.

Keyword arguments: significant_reactions (List): List of reaction IDs with altered flux. model_reactions (List) – List of all reaction IDs considered in the analysis. reaction_to_pathways (Dict): Dictinary mapping reaction ID -> list of pathway names.

Returns: hypergeometric_enrichment_df (pd.DataFrame) – DataFrame with pathway, p-value, counts and adjusted FDR.

distributions_comparison_utils.plot_pathway_enrichment(hypergeometric_enrichment_df: pandas.DataFrame, pval_threshold: float = 0.05, use_fdr: bool = True, font_size: int = 14, width: int = 900, height: int = 600)

Function that takes as input the hypergeometric_pathway_enrichment_df dataframe created with the hypergeometric_test_pathway_enrichment function and plots the enrichment results in a bubble plot. Bubble size stands for pathway size (number of reactions) and reaction colour stands for -log10(p-value)

Keyword arguments: hypergeometric_enrichment_df (pd.DataFrame): Enrichment DataFrame from the hypergeometric_test_pathway_enrichment function. pval_threshold (float): significance p value threshold for filtering. use_fdr (bool): boolean for whether to use ‘fdr’ or ‘pval’ for filtering. font_size (int): font size for plot labels and title. width (int): width of the figure in pixels. height (int): height of the figure in pixels.