Explore and share your scRNAseq clustering results
scClustViz now uses the input data object (Seurat or SingleCellExperiment) plus a new S4 class sCVdata to store all the information needed for the visualization. See ?CalcAllSCV for the single-step setup prior to running the Shiny app. This is not compatible with previously calculated scClustViz results, due to an error in the FDR correction methodology in <v1.0.0
Refactored the DE testing function so that it is made up of modular components, allowing it to be incorporated into existing analysis pipelines. Now there's a function to test DE at a single resolution, so that you can test resolutions as you iterate through them during clustering analysis (see example on webpage - https://baderlab.github.io/scClustViz/#scrnaseq-analysis-pipeline). Each step of the DE testing is also modular, so you can swap in your DE test of choice.
Added the ability to select cells for the set comparison tool based on metadata filters as well as manually selecting cells on the tSNE. Silhouette widths are now a metadata column.
DOI release for publication in F1000Res
Minor bugs in the UI have been fixed Bugs relating to character rather than factor metadata fixed
No major features planned, next step is to rewrite the analysis workflow notebooks to incorporate the package functions, which might result in the addition of more modular analysis functions in the preprocessing steps.
scClustViz is now an R package. There are two functions for loading and DE testing, and one to run the Shiny app from saved data. See the website for more details on usage. Install using devtools:
devtools::install_github("BaderLab/scClustViz")