monkeybread.plot.shortest_distances

monkeybread.plot.shortest_distances(observed_distances, expected_distances=None, show=True, **kwargs)

Plot the distribution of observed distances from each cell in one group (e.g., all T cells) to its closest cell in a second group (e.g., all macrophages) as calculated by monkeybread.calc.shortest_distances().

Optionally, also plot the null distribution of these distances under the null hypothesis that the two cell types do not co-locate. This null distribution must first be computed by monkeybread.stat.shortest_distances().

Parameters:

• observed_distances (DataFrame) – The observed shortest distances, as calculated by monkeybread.calc.shortest_distances().

• expected_distances (Union[ndarray, Tuple[ndarray, float, float], None] (default: None)) – The expected distances under the null hypothesis that the two cell types do not co-locate. Optional to include the distance threshold and p-value, as calculated by monkeybread.stat.shortest_distances().

• show (Optional[bool] (default: True)) – If true, displays the plot(s). If false, returns the Axes instead.

• kwargs – Keyword arguments to pass to seaborn.histplot().

Return type:

Union[Axes, Tuple[Axes, Axes]]

Returns:

: If show = False, returns nothing. Otherwise, returns a single Axes object or a tuple of two Axes objects if expected_distances is provided.

Example

In the left-hand figure, we plot the distribution of distances between each cell in the first group to its closest cell in the second group. The right-hand plot shows the expected distribution under the null hypothesis that cells do not co-locate. The vertical red line denotes the distance threshold used to compute the test-statistic described by the p-value from monkeybread.stat.shortest_distances().