`spateo.segmentation.moran`#

Cell masking using Moran’s I metric.

Adapted from code written by @HailinPan.

Module Contents#

`moranI`(→ Tuple[numpy.ndarray, numpy.ndarray, ...)	Compute Moran's I for cell masking.
`run_moran`(→ numpy.ndarray)	Compute scores using Moran's I method.
`run_moran_and_mask_pixels`(→ numpy.ndarray)	Compute scores using Moran's I method.
`binary_morani_result`(→ numpy.ndarray)	Generate cell mask based on Moran's I.

Compute Moran’s I for cell masking.

Parameters:

X: Numpy array containing (possibly smoothed) UMI counts or binarized values.
kernel: 2D kernel containing weights
mask: If provided, only consider pixels within the mask

Returns:

A 4-element tuple containing (z, c, i, pvalue).

spateo.segmentation.moran.run_moran(X: numpy.ndarray, k: int = 7, p_threshold: float = 0.05, mask: numpy.ndarray | None = None) → numpy.ndarray[source]#

Compute scores using Moran’s I method.

Parameters:

X: Numpy array containing (possibly smoothed) UMI counts or binarized values.
k: Kernel size
p_threshold: P-value threshold. Test. Test Test
mask: If provided, only consider pixels within the mask

Returns:

A 2D Numpy array indicating pixel scores

spateo.segmentation.moran.run_moran_and_mask_pixels(adata: anndata.AnnData, layer: str, k: int = 7, method: str = 'edge-watershed', mk: int = 3, mask: numpy.ndarray | None = None, mask_layer: str | None = None) → numpy.ndarray[source]#

Compute scores using Moran’s I method.

Parameters:

adata: Input Anndata
layer: Layer that contains UMI counts to use
k: Kernel size
method: Method used for generating cell mask based on p value of Moran’s I. ‘edge-watershed’ or ‘otsu’
mk: Kernel size of morphological open and close operations to reduce noise in the mask.
mask: If provided, only consider pixels within the mask
mask_layer: Layer to save the final mask. Defaults to {layer}_mask.

Returns:

A boolean mask.

spateo.segmentation.moran.binary_morani_result(c: numpy.ndarray, p: numpy.ndarray, pvalue_cutoff: float = None, method: str = 'edge-watershed', c_cutoff: float = None, tissue_mask: numpy.ndarray | None = None) → numpy.ndarray[source]#: Generate cell mask based on Moran’s I.