crested.pl.design.step_contribution_scores

crested.pl.design.step_contribution_scores#

crested.pl.design.step_contribution_scores(intermediate, scores_all, seqs_one_hot_all, sequence_labels=None, class_labels=None, zoom_n_bases=None, ylim=None, global_ylim='per_plot', method=None, highlight_kws=None, show=True, labels='deprecated', **kwargs)#

Visualize enhancer design stepwise contribution scores.

Contribution scores can be calculated using the contribution_scores() method.

Parameters:

  • intermediate (list[dict] | dict) – Intermediate output from enhancer design when return_intermediate is True. Specifically, a single dict or list of dicts (one per designed sequence), with dict entry ‘changes’ to denote the changed position per step.

  • scores_all (list[ndarray] | ndarray) – A list of contribution scores arrays for each designed sequence, of shape [(seq_steps, n_classes, n_bases, n_features), …], like from contribution_scores().

  • seqs_one_hot_all (list[ndarray] | ndarray) – A list of one-hot encoded corresponding sequence arrays of shape [(seq_steps, n_bases, n_features), …], like from derive_intermediate_sequences().

  • sequence_labels (list | None (default: None)) – List of sequence labels ot add to the plot. Should have the same length as the number of designed sequences.

  • class_labels (list | None (default: None)) – List of class labels to add to the plot. Should have the same length as the number of classes.

  • zoom_n_bases (int | None (default: None)) – Number of center bases to zoom in on. Default is None (no zooming).

  • ylim (tuple[float, float] | None (default: None)) – Y-axis limits, ignored if global_ylim is set. Default is None.

  • global_ylim (Optional[Literal['all', 'per_design', 'per_plot']] (default: 'per_plot')) – Used to set global y-axis limits across explanations. Can be one of ‘all’, ‘per_design’, ‘per_plot’ or None. Default is ‘per_plot’ ‘all’ makes the y-axis limit same across all of the explanations. ‘per_design’ makes the y-axis limit same across all of the steps and classes of a single designed sequence. ‘per_plot’ makes y-axis limits same across all the steps but not the classes of a single designed sequence. If None, each explanation has its y-axis limits separately selected.

  • method (Optional[Literal['mutagenesis', 'mutagenesis_letters']] (default: None)) – Default is None (for gradient-based contributions). If plotting mutagenesis values, set to 'mutagenesis_letters' (to visualize average effects as letters) or mutagenesis (to visualize in a legacy way).

  • highlight_kws (dict | None (default: None)) – Keywords to use for plotting changed basepairs with axvspan(). Default is {‘edgecolor’: “red”, ‘facecolor’: “none”, ‘linewidth’ :0.5}

  • show (bool (default: True)) – Whether to show all plots or return the (list of) figure and axes instead.

  • width – Width of each created figure. Default is 50.

  • height – Height of each created figure. Default is 2*`n_seqs`.

  • sharex – Whether to share the x axes of the created subplots within each figure. Default is False.

  • sharey – Whether to share the y axes of the created subplots within each figure. Default is False.

  • save_path – If provided, where to save the figure(s). If more than one plot is created (like with multiple sequences and/or multiple classes), will append _1/_2/… to the filename.

  • kwargs – Additional arguments passed to contribution_scores() to control contribution score settings and on to render_plot() to control the final plot output. Please see contribution_scores() and render_plot() for details. Custom defaults for enhancer_design_steps_contribution_scores: suptitle_fontsize=26, tight_rect=[0, 0, 1, 0.98].

Return type:

tuple[Figure, list[Axes]] | list[tuple[Figure, list[Axes]]] | None

Returns:

If show=False, a (fig, axs) tuple (if plotting one sequence and one class), or a list of (fig, axs) tuples (if plotting multiple sequences and/or classes).

See also

crested.pl.render_plot, crested.pl.explain.contribution_scores

Example

>>> crested.pl.design.step_contribution_scores(
...     intermediate_results,
...     scores,
...     one_hot_encoded_sequences,
...     class_labels=["L5ET"],
...     highlight_kws={'facecolor': 'green', 'edgecolor': 'green', 'alpha': 0.1},
... )
../../_images/design_step_contribution_scores.png
Contents