SCENE is a analytic tool for dissecting cellular cohabitation using single-cell and bulk RNA-seq data.

The SCENE package requires computers with enough RAM to support the operations defined by a user. Applying SCENE on 5000 samples (comprising single cells and bulk samples) requires about 12 GB of RAM and about 40 minutes to complete all calculation tasks on a workstation with 96 CPU threads(Intel(R) Xeon(R) Gold 6240R CPU @ 2.40GHz). For larger cohorts, we recommend running the following functions on a high performance computer cluster and visualizing the results locally: preprocessing2, determineK_runNMF, runcNMF, and identifyConsensusProgram3.

The SCENE development version is developed and tested on Linux operating systems. The developmental version of the package has been tested on the following systems:

• Linux: Ubuntu 20.04
• MacOS: Sonoma 14.0 (23A344)

Before setting up the SCENE package, users should have R version 4.0.5 or higher installed.

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("lijxug/SCENE")

Installation typically takes several minutes, depending on the installation of dependencies.

Full documentation is being drafting and will be online soon. A demo case is currently available at Code Ocean (Capsule DOI: 10.24433/CO.4624314.v1). All the heavy calculations in this capsule have been performed previously. Running this capsule online takes about 20 minutes to establish the environment and 5 minutes to print the figures.

• version 1.0: First released version

The code is released under the GNU GPL-3 License.