Scaling of conditions

[roanvanscheppingen]

Hi,

I've been working on analysis of spatial data. It consists out of multiple TMAs of tissue on a single slide, consisting out of 2 conditions. I integrate the data based on TMA, since I view them as technical replicates (although still on the same slide).

However, upon inspecting Z-scores for all genes in the set, I see that condition A always has lower Z-scores than condition B. This is also reflected in the 'counts' and 'data' slot, with condition A being persistently lower.

1. When I use Subset <- ScaleData(Subset, vars.to.regress = c("nCount_RNA"), split.by = "Condition"), should split.by be used here?
2. Lets say Condition A's mean after integration is 100, and that of condition B is 125, should I take the "data" layer of all cells from condition A and multiply by 1.25 as a sort of library size scaling?

Thank you for your time!