scRNAseq Cell-Cell Communication Branch Tutorial

This tutorial will assist users with analyzing scRNAseq data using CellChat. R is required for this analysis. Rather than using miniconda, this tutorial uses the uwot UMAP method.

Data Download

The data required for the analysis done in this tutorial is included in the data folder of this directory.

Running cellchat.rmd

1. Before performing any analysis users must make sure cellchat and all dependencies are installed:

   devtools::install_github("sqjin/CellChat")
   

   Other dependencies:

   • Install NMF (>= 0.23.0) using install.packages('NMF').
   • Install circlize (>= 0.4.12) using devtools::install_github("jokergoo/circlize") .
   • Install ComplexHeatmap using devtools::install_github("jokergoo/ComplexHeatmap").

2. It is recommended to run cellchat.rmd chunk-by-chunk to ensure all dependencies are properly installed.

Modify the internal netClustering function

This workflow was written using the old way of implementing parallel processing, causing an issue when the software package version was updated. To check the netClustering function follow these steps:

1. Type the function name and R will print out the source code of the function.
2. You can find if the parallel function was enabled by default
3. Set the parameter to false in the netClustering function so the use of the parallel processing package is skipped.

do.parallel = FALSE

Outputs

Running cellchat.rmd will generate an HTML file. This file will contain the following figures:

• Pie charts
• Circle plots
• Heat maps
• Chord diagrams
• Bubble plots
• Violin/ dot plots
• Scatter plots
• River plots