Mudata format for MOFA use.

[soniyalama]

Hi,
thanks for your tool.
I am interested in the interactions between different cell types, and how they differ across conditions (X and Y), which tool do you think is best to use, in order to do so?

I was first thinking of just subsetting my original adata subset in the conditions X and a subset condition Y. And then running LIANA+.
However, I also saw your option of using MOFA, but is that only possible for Mudata, or also for anndata? I'm working with anndata and was considering trying out the MOFA, but then I saw the requirement of needing Mudata.

Thanks in advance.

[dbdimitrov]

Hi @soniyalama,

Thanks for the issue. However, I'm not 100% sure if I understand your objective.

It might help if you provide some details - e.g. what type of data, which liana+ method/tutorial are you referring to, etc.

Daniel

[soniyalama]

My apologies for not being as clear. I'm working with single-cell data. my main goal is to check how the interactions between different cell types differ between two conditions (treatment vs non-treatment). I was thinking if your MOFAcellular - Multicellular Factor Analysis vignette would be able to help me with that, or if it's better to stick with the Steady-state Ligand-Receptor inference?
Thanks again.

[dbdimitrov]

Hi @soniyalama, for cross conditions, mofa or alternatively tensor-cell2cell would fit best indeed 🙂