About spatial proximity test

[Pedramto89]

Hi there

I did spatial proximity test and I have some questions.

1- How would you explain cell to cell pairs that sound to be close to each other based on spatial proximity test but not visually? I mean, I did the test and it seems some cells sound to co-located but when I look at the spatial images I have , I see other co-located cells which are not totally aligned with spatial proximity test results.

2- It appears that the spatial proximity test is built on both gene expression and spatial coordinates which are being calculated via KNN. I am wondering is there anyway to do the spatial proximity test only based on coordinates without including gene expression?

3- Also, I transferred some of my data from Seurat Obj to create the Giotto Obj and then do the spatial proximity. When I see the spatial plot on Giotto, it sounds there are some discrepancies between what the spatial plot shows on Seurat and what Giotto shows. What do you guess about the root of the problem?

Best