crested.pl.region.bar

Contents

crested.pl.region.bar#

crested.pl.region.bar(data, region=None, model_names=None, classes=None, log_transform=False, pred_color='tab:blue', truth_color='green', plot_kws=None, ax=None, adata='deprecated', target='deprecated', **kwargs)#

Barplot of ground truths and/or predictions for a specific region in your data.

Can plot either a region from an AnnData (data=adata, region=region) or manual prediction (data=pred, classes=adata.obs_names). When plotting from AnnData, can plot the ground truth and any available predictions.

Parameters:
  • data (AnnData | ndarray) – AnnData object containing the genomic data in var (requiring the region argument), or single prediction numpy array (requiring the classes argument).

  • region (str | None (default: None)) – Region name from the AnnData, generally in format ‘chr:start-end’. Required if providing an AnnData object.

  • model_names (str | list[str] | None (default: None)) – Source of the values to plot, as a name or list of names. Can be ‘X’/’truth’/’groundtruth’ for the ground truth from adata.X, or the name of prediction layers in adata.layers. If None, plots the ground truth and all layers in the AnnData. Disregarded if plotting a single prediction.

  • classes (list[str] | None (default: None)) – The class names to use. Required if providing a single prediction. If using an AnnData object, read from there by default.

  • log_transform (bool (default: False)) – Whether to apply a log1p transformation to the data.

  • pred_color (str (default: 'tab:blue')) – Plot color of any prediction barplot.

  • truth_color (str (default: 'green')) – Plot color of any ground truth barplot.

  • plot_kws (dict | None (default: None)) – Extra keyword arguments passed to bar(). Defaults: 'alpha': 0.8.

  • ax (Axes | None (default: None)) – Axis to plot values on. If not supplied, creates a figure from scratch.

  • width – Width of the newly created figure if ax=None. Default is 18.

  • height – Height of the newly created figure if ax=None. Default is 4*n_model_names.

  • kwargs – Additional arguments passed to render_plot() to control the final plot output. Please see render_plot() for details. Custom defaults for region: title='Ground truth'/target, 'color'=['tab:blue'`(for predictions)/’green’(for ground truth)]`, suptitle=region, ylabel="Ground truth"/"Prediction, grid='y'.

Return type:

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

Example

>>> crested.pl.region.bar(
...     adata,
...     region='chr1:3093998-3096112',
...     model_names="Base model",
...     width=20,
...     height=3
... )
../../_images/region_bar_single.png
>>> crested.pl.region.bar(
...     adata,
...     region="chr1:3093998-3096112",
...     pred_color="royalblue",
...     width=20,
... )
../../_images/region_bar_all.png
>>> crested.pl.region.bar(
...     pred,
...     classes=adata.obs_names,
...     title="Designed Micro-PVM enhancer",
... )
../../_images/region_bar_pred.png