9 Normalising CosMx data”
Data normalisation is a tricky step. There are multiple opinions about it in the feild. Tools such as SpaNorm
have been specifically developed for normalising image-based spatial transcriptomics data. However, for this data set, I foudn out that LogNormalise
actually work well. So that is what Im doing below. But depending on the data structure, LogNormalise
may not work for your data.
merged <- NormalizeData(
merged,
normalization.method = "LogNormalize",
verbose = TRUE,
scale.factor = 10000,
assay = "RNA")
Challenge: Explore the functionality of SpaNorm
together.