Skip to content
/ spade Public

SPADE: Spanning Tree Progression of Density Normalized Events

License

Notifications You must be signed in to change notification settings

nolanlab/spade

Repository files navigation

SPADE: Spanning Tree Progression of Density Normalized Events

SPADE is a visualization and analysis tool for high-dimensional flow cytometry data. SPADE is implemented as an R package that takes in FCS files and settings, and outputs graph files (GML), PDFs of the graphs and FCS files with the "cluster" column appended. Please see the project homepage at cytospade.org or the wiki for the primary documentation.

We no longer support the Cytoscape GUI. An interactive viewer is available from www.cytobank.org. The PDF files can be edited (e.g. graphs can be rearranged) with software such as Adobe Illustrator. For more complex usages, many languages are able to read GML files, including R, Mathematica and Python.

User Setup

We no longer support distribution via CRAN or Bioconductor. The instructions below install directly from Github and will ensure you have the latest release.

  1. If you do not have version 3.0 or later of R, install it (OSX, Windows, Linux). Unless you have a compelling reason to do otherwise, we suggest the 64-bit version of R.

  2. Install the devtools, Rclusterpp (latest), the bioconductor installer and SPADE packages:

     R> install.packages("devtools")
     R> library(devtools)
     R> devtools::install_github("nolanlab/Rclusterpp")
     R> source("http://bioconductor.org/biocLite.R")
     R> devtools::install_github("nolanlab/spade")
    

(The version of Rclusterpp in CRAN (0.2.3) is out of date; the above installs the latest (i.e. >= 0.2.4).)

  1. Test the installation by typing at the R prompt:

     R> library(spade)
    

    You should see output like the following:

     R> library('spade')
     Loading required package: igraph
     Loading required package: Rclusterpp
     Loading required package: Rcpp
     Loading required package: RcppEigen
    

Usage

Check out the example usage wiki page. Once you are familiar with the workflow, you can edit the file R/inst/runSPADE.R to setup your analysis, then run the file. For additional documentation about the R package, you can view the package vignette with vignette("SPADE") at the R prompt. Additionally all of the functions in the SPADE R package are documented; view their manual pages with ?<function>, e.g., ?SPADE.driver, at the R prompt.

Developer Setup

Please refer to the wiki pages.


Citations

SPADE was developed in the Plevritis and Nolan Labs at Stanford University, and is described in the following publications: