search

satijalab / seurat

R toolkit for single cell genomics
http://www.satijalab.org/seurat
Other
2.31k stars 919 forks source link

Spatial transcriptomics, Harmony & normalisation #6028

Closed JJBio closed 1 year ago

JJBio commented 2 years ago

I am analysing multiple Visium slides and have a grave batch effect. As suggested by 10x, I would like to correct for the batch effect using Harmony. However, SCtransform and Harmony (used according to this workflow), did not give me good results. In contrast, using the 'old' normalisation (lognormalize) works much better (at least the results are biologically more sensible). Additionally, I would like use my scRNA-seq dataset for integration, which is also processed with lognormalize. Furthermore, I am avoiding this issue, correct? So, the final corrected values are not comparable across samples. Nevertheless, the vignette states:

As a result, standard approaches (such as the LogNormalize() function), which force each data point to have the same underlying ‘size’ after normalization, can be problematic.

Hence, my final question is how problematic is using lognormalize on spatial data? Would you absolutely not recommend it?

rsatija commented 1 year ago

While we use SCTransform in the vignettes, it is certainly not problematic to use LogNormalize when doing spatial workflows