Code accompanying the CytoMDS paper

Code and raw data to reproduce the results in the CytoMDS article.

User's instruction:

Set-up

• do a git clone of this github repo

• (if you are using RStudio) create a R project pointing to the root directory

• download and copy the raw data :

  • for the HBV Chronic Mouse dataset, download the corresponding zenodo repository, and copy the rawData and compensation directories into the ./data/HBV_chronic_mouse/ subdirectory of the current repository.
  • for the ImmunoSenescence human PBMC dataset, a controlled access solution is currently being set-up by GSK, the dataset owner (update in a future version of the article).
  • for the Krieg Anti PD1 dataset, no need to download anything, since download will happen automatically at execution of the R scripts thanks to the HDCytoData package.

• if not done yet, install the needed packages: CytoPipeline and CytoPipelineUtils for pre-processing, CytoMDS, CytoPipelineGUI and some other packages for the generation of figures, as shown in the comment code below:

# if (!require("BiocManager", quietly = TRUE))
#     install.packages("BiocManager")
# 
# BiocManager::install("CytoPipeline")
#
# if (!require("devtools", quietly = TRUE))
#     install.packages("devtools")
# devtools::install_github("UCLouvain-CBIO/CytoPipelineUtils")
#
# BiocManager::install("CytoMDS")
# BiocManager::install("CytoPipelineGUI")
# BiocManager::install(c("patchwork", "Polychrome", "purrr"))
# 
# BiocManager::install("sva")
# devtools::install_github("biosurf/cyCombine")

IMPORTANT REMARKS:

• the pre-processing part requires CytoPipeline version >=1.5.1 .
• the MDS plot part requires CytoMDS version >=1.3.3, which is in Bioconductor 3.21, i.e. the development version at the time of writing.

Generating the CytoPipeline caches for pre-processing

Execute the following two R statements :

source("./R/HBV_chronic_mouse_CytoPipeline_Preprocessing.R")
source("./R/ImmS_CytoPipeline_Preprocessing.R")

Note that each statement is likely to take up to a few minutes.

Producing the article plots

Execute the following R statements:

source("./R/Plots_Common.R")
source("./R/HBV_chronic_mouse_CytoMDS.R")
source("./R/ImmS_CytoMDS.R")
source("./R/Krieg_Anti_PD_1_CytoMDS.R")
source("./R/ImmS_ComputationalStudy.R")

The plots are generated in directory ./plots/.