In Xenium (or other panel-based methods) we only sample a small fraction of the genes (e.g. 5%), so we'll be missing a lot of the receptor-ligand pairs, making an analysis like cellchat extremely unreliable.
There might be some papers to support this?
In Xenium (or other panel-based methods) we only sample a small fraction of the genes (e.g. 5%), so we'll be missing a lot of the receptor-ligand pairs, making an analysis like cellchat extremely unreliable.
There might be some papers to support this?