Update compositional.md

bcbio · Jan 31, 2025 · 4049230 · 4049230
1 parent fc15b32
commit 4049230
Showing 1 changed file with 8 additions and 3 deletions.
diff --git a/inst/templates/singlecell/compositional/compositional.md b/inst/templates/singlecell/compositional/compositional.md
@@ -1,6 +1,11 @@
-Meeta and I have examined MiloR, propeller, and sccomp. MiloR transforms to neighborhoods so is not comparable. propeller and sccomp give similar results but with some striking differences. Best practice may be to run both and see. The most conservative approach would be to take the intersect.
-
+Using the brown fat scRNA-seq dataset (TN vs cold7), we ran both propeller and sccomp to evaluate the results. For a majority of the celltypes there as concordance showing that proprtions of cells were significantly different between conditions. Howver, for celltypes identified by only one of the methods the result was not always convincing. Best practice for the template may be to run both methods and compare results. The most conservative approach would be to take the intersect of the two methods.
 
 <p align="center">
 <img src="comp.png" width="500">
-</p>
+</p>
+
+
+Another method, currently being used in the training material, is MiloR. This tool transforms to neighborhoods so is not directly comparable to propeller and sccomp.
+
+
+