Handling Multiple Cells Sharing the Same Coordinates in Squidpy Neighborhood Analysis

[aboutfanfan]

Hi,

I am working with spatial transcriptomics data (Stereo-seq bin50) that has been processed using CytoSPACE. After aligning with single-cell data, each spatial spot has been decomposed into multiple cells of different types. As a result, multiple cells now share the same xy coordinates, but they represent distinct cell types, creating a new dataset based on cell types rather than the original spatial spots.

I would like to perform neighborhood analysis using Squidpy, but I am unsure how to handle the fact that multiple cells share the same xy coordinates. Here are my main concerns:

1. Should I add slight random offsets to the shared coordinates to differentiate them spatially?
2. Is there a way to aggregate these cells or their types for neighborhood analysis (e.g., by creating a weighted matrix or composite types)?
3. What would be the recommended practice in Squidpy for this scenario to ensure biologically meaningful results?

Any suggestions or best practices would be greatly appreciated!

Thank you for your help!