adding a note about using SCT for findmarkers

lessons/09_merged_SC_marker_identification.md

@@ -111,7 +111,12 @@ Before we start our marker identification we will explicitly set our default ass
 DefaultAssay(seurat_integrated) <- "RNA"
 ```
 
-> _**NOTE:** The default assay should have already been `RNA`, because we set it up in the previous clustering quality control lesson. But we encourage you to run this line of code above to be absolutely sure in case the active slot was changed somewhere upstream in your analysis. Note that the raw and normalized counts are stored in the `counts` and `data` slots of `RNA` assay. By default, the functions for finding markers will use normalized data._
+The default assay should have already been `RNA`, because we set it up in the previous clustering quality control lesson. But we encourage you to run this line of code above to be absolutely sure in case the active slot was changed somewhere upstream in your analysis. 
+
+> #### Why don't we use SCT normalized data?
+> Note that the raw and normalized counts are stored in the `counts` and `data` slots of `RNA` assay, respectively. By default, the functions for finding markers will use normalized data if RNA is the DefaultAssay. The number of features in the `RNA` assay corresponds to all genes in our dataset.
+>
+> Now if we consider the `SCT` assay, functions for finding markers would use the `scale.data` slot which is the pearson residuals that come out of regularized NB regression. Differential expression on these values can be difficult interpret. Additionally, only the variable features are represented in this assay and so we may not have data for some of our marker genes.
 
 The function `FindConservedMarkers()`, has the following structure: