Tools for the analysis of .fcs data sets
Enter the following command within an R session
devtools::install_github("jsim91/FCSimple")
- flowCore
- ncdfFlow
- flowWorkspace
- uwot
- Rtsne
- FNN
- Matrix
- igraph
- FlowSOM
- Rphenograph
- ComplexHeatmap
- circlize
- CATALYST
- ggpubr
- ggplot2
- shadowtext
- future
- future.apply
- RANN
- shiny
- shinythemes
- scales
- cytoMEM
- dbscan
- cyCombine
The dimension reduction and clustering steps do offer methods to run the calculations through Python, however no knowledge of the Python language is required. These functions will call Python in the background and results will be transferred to the R environment. The reticulate package is not required. Users may edit the included python scripts if they want to alter the default Python behavior. Scripts are located at: main/inst/python. To take advantage of Python-supported methods, these items are required:
- Python # Check the "Add to Path" box during installation. The Python installation should come with pip.
- FlowKit # *** necessary for hyperlog transformation of data when using fcs_join() ***
- umap # umap has its own set of dependencies, see link
- pynndescent # Required to use this package's UMAP-by-Python functionality
- leidenalg # leiden in Python
- scipy
- numpy
- pandas
- igraph # louvain in Python
- openTSNE # tSNE in Python
- GIT # GIT clustering algorithm
This package requires the user to initiate a list object like this:
include_files <- list.files(path = "fcs_file_directory", pattern = ".fcs", full.names = TRUE) # Point to the directory where the FCS file(s) of interest are located. It's recommended to use full names.
my_object <- fcs_join(files = include_files) # This function takes the listed files as input and will initialize the list object that will store all subsequent analyses.
Subsequent functions will take the object created with function
fcs_join()
as input. Analysis outputs will be joined to this object as new list entries. It's highly recommended to set the transform parameters for all relevant features in Flowjo then pass the workspace diagnostics output (as .txt file) to fcs_join() using the flowjo_diagnostics_file argument. This is how you can get the workspace diagnostics output in Flowjo:
Tested in and developed with the Windows 10/11 operating system.