Thank developer for the detailed tutorial.

The best place for this would be the forum, but my post there seem to be automatically hidden or deleted by the system. So try asking the question here.

Should we apply score_genes or score_genes_cell_cycle to the normalized matrix or to the scaled matrix?
From the documentation, it is shown that it selects genes with similar expression as controls and calculates the difference between the two sets.
In this way it seems that it should be applied to the normalized matrix.
But from the tutorial, this function is applied to the scaled matrix. It also seems from the source code that the average expression stored in the adata.var was not used to select the control genes.

In the corresponding function of Seurat, AddModuleScore, this function is applied to the normalized matrix.

I would like to know what type of matrix we should apply score_genes to in the ideal case.
Although they are not very different in the example data (nestorawa_forcellcycle_expressionMatrix.txt).

I would love to hear your suggestions on this.
Thanks very much.