crested.pl.region.scatter

crested.pl.region.scatter(adata, region, model_names=None, log_transform=False, annotate_top=5, top_method='sum', square=True, identity_line=False, plot_kws=None, annotate_kws=None, ax=None, **kwargs)

Scatterplot of ground truths and predictions for a specific region in your data, comparing different classes.

Parameters:

• adata (AnnData) – AnnData object containing the genomic data in var.

• region (str) – Region name from the AnnData, generally in format ‘chr:start-end’.

• model_names (str | list[str] | None (default: None)) – The target to plot the distribution for, either None (to plot all) or one/multiple names of a prediction layer in adata.layers.

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

• annotate_top (int | None (default: 5)) – Number of top classes to annotate with their class names. If None, does not annotate any.

• top_method (Literal['sum', 'truth', 'pred'] (default: 'sum')) – How to determine the top n classes. “truth” or “pred” use highest on either, “sum” takes highest sum of both.

• square (bool (default: True)) – Whether to force the plots to be square, have equal aspect ratios, and equal shared axis ranges.

• identity_line (bool (default: False)) – Whether to plot a y=x line denoting perfect correlation.

• plot_kws (dict | None (default: None)) – Extra keyword arguments passed to scatter(). Defaults: 's': 10, 'color': 'white', 'edgecolor': 'black'.

• annotate_kws (dict | None (default: None)) – Extra keyword arguments passed to annotate(). Defaults: 'xytext': (5, 0), 'textcoords': 'offset points'

• 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 7*n_models.

• height – Height of the newly created figure if ax=None. Default is 7.

• sharex – Whether to share the x axes of the created plots. Default is False. Setting square=True does equalize limits even if sharex=False

• sharey – Whether to share the y axes of the created plots. Default is False. Setting square=True does equalize limits even if sharey=False.

• kwargs – Additional arguments passed to render_plot() to control the final plot output. Please see render_plot() for details. Custom defaults for region: xlabel="Ground truth", ylabel="Prediction", `title="region", layout="compressed", title="{region}-{model_name}"`(1 model)/`model_name`(>1 model), `suptitle=region (>1 model).

Return type:

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

See also

crested.pl.render_plot, crested.pl.region.bar

Example

>>> crested.pl.region.scatter(
...     adata,
...     "chr1:3111367-3113481",
...     model_names='Base model',
...     log_transform=True,
...     identity_line=True,
... )