Hi,
I have previously used scCODA of pertpy for compositional analysis on single-cell data.
However, I am unsure whether the same approach can be applied to annotated spatial transcriptomics spots. specifically, if spatial spots have already been annotated, is it appropriate to treat these spots analogously to samples in scCODA and perform compositional analysis across conditions?
At first glance I don’t see why you shouldn’t be able to do so.
Let’s say instead of spatially sequencing your samples, you’d somehow dissociate and sort it into cell types (or whatever you annotate your spots with). If now you can somehow quantify what fraction of cells from your tissue went into each category (per sample), then you could test for enrichment of certain categories in samples of a certain condition. Whatever that hyophetical experiment could answer, scCODA should be able to tell you based on your spatial dataset.
Unless my answer tells you that I misunderstood your question, or somebody else chimes in and corrects me, I’d suggest to give it a try and think of some positive/negative controls that you could use to increase your confidence in the results.