cellrank.pl.aggregate_fate_probabilities

cellrank.pl.aggregate_fate_probabilities(adata, mode='paga_pie', backward=False, lineages=None, cluster_key='clusters', clusters=None, basis=None, cbar=True, ncols=None, sharey=False, fmt='0.2f', xrot=90, legend_kwargs=mappingproxy({'loc': 'best'}), figsize=None, dpi=None, save=None, **kwargs)

Plot aggregate lineage probabilities at a cluster level.

This can be used to investigate how likely a certain cluster is to go to the terminal states, or in turn to have descended from the initial states. For mode ‘paga’ and ‘paga_pie’, we use PAGA [Wolf et al., 2019].

Parameters:

• adata (AnnData) – Annotated data object.

• mode (Literal['bar', 'paga', 'paga_pie', 'violin', 'heatmap', 'clustermap']) –

  Type of plot to show. Valid options are:

  • 'bar' - barplot, one panel per cluster. The whiskers correspond to the standard error of the mean.

  • 'paga' - paga(), one per initial or terminal state, colored in by fate.

  • 'paga_pie' - paga() with pie charts indicating aggregated fates.

  • 'violin' - violin plots, one per initial or terminal state.

  • 'heatmap' - a heatmap, showing average fates per cluster.

  • 'clustermap' - same as a heatmap, but with a dendrogram.

• backward (bool) – Direction of the process.

• lineages (str | Sequence[str] | None) – Lineages for which to visualize the fate probabilities. If None, use all lineages.

• cluster_key (str | None) – Key in obs containing the clusters.

• clusters (str | Sequence[str] | None) – Clusters to visualize. If None, all clusters will be plotted.

• basis (str | None) – Basis for scatterplot to use when mode = 'paga_pie'. If None, don’t show the scatterplot.

• cbar (bool) – Whether to show colorbar when mode = 'paga_pie'.

• ncols (int | None) – Number of columns when mode = 'bar' or mode = 'paga'.

• sharey (bool) – Whether to share y-axis when mode = 'bar'.

• fmt (str) – Format when using mode = 'heatmap' or mode = 'clustermap'.

• xrot (float) – Rotation of the labels on the x-axis.

• figsize (tuple[float, float] | None) – Size of the figure.

• legend_kwargs (Mapping[str, Any]) – Keyword arguments for legend(). For mode = 'paga_pie' and basis = '...', this controls the placement of the fate probabilities legend.

• figsize – Size of the figure.

• dpi (int | None) – Dots per inch.

• save (str | Path | None) – Filename where to save the plot.

• kwargs (Any) – Keyword arguments for paga(), violin() or bar(), depending on the mode.

Return type:

None

Returns:

: Nothing, just plots the figure. Optionally saves it based on save.

Notes

For mode = 'paga_pie', directed edges are shown only when adata.uns['paga']['transitions_confidence'] is present. This key is produced by scVelo’s PAGA extension (paga()), which augments the standard scanpy PAGA graph with directed transition confidences. When only scanpy’s paga() has been run, edges are undirected.