Code repository for data processing and plotting included in our study characterizing donor Treg in the prophylactic treatment of acute GvHD.
Donor regulatory T cells rapidly adapt to recipient tissues to control murine acute graft-versus-host disease.
The repository is structured based on data packages that were analysed (mostly) independent from each other. This includes:
- Analysis of bulk RNA sequencing data (bulkRNAseq_analysis folder)
- Analysis of bulk TCR sequencing data (bulkTCRseq_analyses folder)
- Analysis of single cell RNA sequencing data (scRNAseq_analysis folder)
- Analysis of single cell TCR sequencing data (scTCRseq_analysis folder)
- Plotting of additional data (FACS, etc.) (Rscripts & data folder)
- Perl scripts used for processing or plotting (perlScripts folder)
This directory contains data and code for the analyses underlying the bulkRNAseq experiments of re-isolated Treg and the corresponding expanded donor Treg cells.
The following figures are covered:
- Figure 1 e, f
- Supplementary Figure 1 d
- Supplementary Figure 2 e
- Figure 3 a - c, e, f
- Supplementary Figure 3 c - m
- Figure 4 a - d
- Supplementary Figure 4 a - d
Before running the first .Rmd file for the initial setup (bulkRNAseq_analyses_initial_setup.Rmd), all relevant ...ReadsPerGene.txt files have to be retrieved from GEO and locally stored in a directory called "RPGDIR" within the "bulkRNAseq_analyses" directory within the project directory "tregprophylaxispaper" (/.../tregprophylaxispaper/bulkRNAseq_analyses/RPGDIR). All relevant .rds files for annotation, metadata and read counts are supplied as well. However, in case you want to reproduce metadata and read count data objects, above steps are recommended.
All .Rmd files were successfully run on the system build supplied in the respective SessionInfo.txt files. RStudio is recommended for running the markdown files. Using knitr, a html report can be generated. The code will result in the figure(s) specified. Make sure your project working directory is set two levels above the "RScripts" directory within the "bulkRNAseq_analyses" directory. When using git to clone the repository, this should be the case by default and the "tregprophylaxispaper" directory should be specified as your project working directory.
A certain order of running the markdown files is recommended as some analyses rely on previous analyses' results, e.g. "Figure3_TregDGEScatter_GSEA_cvnc.pc_BMT.Rmd" requires results from "Figure3_TregDGEScatter_GSEA_Heatmap_Metascape_cvnc.b.Cd62Lneg.Rmd" for a GSEA. Hereafter, the recommended running order is lined out.
- bulkRNAseq_analyses_initial_setup.Rmd
- Figure1_TregDGEScatter_GSEA_avp_invitro
- Figure1_TregSigHeatmap_fresh.allo.poly.Rmd
- FigureS2e_additionalBarplots_pro.don.Rmd
- Figure3_TregDGEScatter_GSEA_Heatmap_Metascape_cvnc.b.Cd62Lneg.Rmd
- Figure3_GSEA_prc.pro_avp.c.l.s.indiv.Rmd
- Figure3_TregDGEScatter_GSEA_cvnc.pc_BMT.Rmd
- Figure3and4_UMAP_Heatmaps_GSEA_Barplots_pro.prc.bas.don.Rmd
- Figure4_TregDGEScatter_GSEA_pvc.cls.Rmd
Of note, for scripts 2., 3., 4., 5., 6., the order is arbitrary. As long as they have been run after 1. and before 7.. Also, for .Rmd files 8., 9. can be run as soon as 1. has been executed.
The code for processing and analysis of the TCR data is provided in several bash scripts, which either call perl or R scripts for analyses and plotting. The following figures are covered:
- Figure 1 c
- Supplementary Figure 1 a
- Supplementary Figure 1 c
- Figure 5 a - f
- Supplementary Figure 5 a - l, n, o
The code for processing and analysis of the scRNA data and scTCR data mapped to scRNA data is provided in a bash script, which either call to .R or .Rmd scripts for analyses, or directly by .Rmd scripts. The following figures are covered:
- Figure 6 a - d
- Figure 6 g - h
- Supplementary Figure 6 a - f
- Figure 7 a - f
- Supplementary Figure 7 a - b
- cellranger_doublets.csv
- doubletFinder.R
- elbowPoint.R
- doubletFinder.html (example plus session info)
- doubletFinder.Rmd
- ElbPlxy.txt
- environment_kb-python_0.26.4.yml (conda environment)
- kb_tools_RNAVelocityanalysis.ipynb (kb_tools_RNAVelocityanalysis.py)
- velofigure.sh (depends on:
scVelo.py)
The code for processing and analysis of the TCR data is provided in a bash script, which either call perl or R scripts for analyses. The following figures are covered:
- Figure 6 e - i
- Supplementary Figure 6 c - f
- Figure 7 e, f
Contains R markdown scripts (*.rmd) that are rendered locally using data provided in the data folder as well as corresponding Session Infos. The following figures are covered:
- Figure 1 b, d
- Figure 2 b - e
- Supplementary Figure 1 c
- Supplementary Figure 2 b, c
Contains data for Figures 1b,d & 2b-e & Supplemental Figures 2b,c.
Perl script to map TCR repertoires derived from our RACE-PCR approach. Sequencing data (usually MiSeq 300bp PE) are initially assembled (paired) using the software PEAR. UMI should be standard in our protocol, but the pipeline can also be used for non-barcoded RNAs (when -UMI option is omitted). MIGs are called using MIGEC before mapping with MIXCR (version 3). We usually run two libraries for each sample (replicate) and work with the combined repertoire from both technical replicate runs. Individual runs are kept for quality controls. The pipeline is setup for our server and deploys folders for each experiment on the misc/data volume under processedData//mapping/RNA/RepSeq/MIXCRv3/. ***Dependencies:*** PEAR (v0.9.11), MIGEC (v1.2.9), MIXCR (v3.0.13), R version 4.0.3 (rbioc_3-12).
Many analyses are done in R, but for two types of presentations (Circos plot of clonotype overlap, and barycentric distributions of clonotype weights), we have a set of Perl scripts that perform preprocessing/reformating of the data (bulkTCRparseImmunArch.pl, bulkTCRcreateJoinedCountTable.pl) and one each for the plotting (bulkTCRCirculizeFreq4Samples.pl, bulkTCRtriangle.pl)
Perl script to reformat the data table generated by the immunoarch R package to fit into our count table joining script. Perl script generating joined clonotype tables from several samples. Accepts files generated by the bulkTCRparseImmunArch.pl script. Perl script that generates a circos plot for clonotype overlaps in exactly four samples. This is tailored for our three animal/one donor design and shows how much overlap there is between donors and recipient organs. ***Dependencies:*** R version 4.0.3 (rbioc_3-12), R packages: circlize, RColorBrewer, gridExtra, grid Perl script that calculates weights of clonotype counts for three sample comparisons and generates a bubble plot using barycentric coordinates by generating and calling a R script. This is a nice way to demonstrate the contribution of each of the three samples to the top clonotypes in the comparison. It accepts clonotypeTable files generated with the bulkTCRcreateJoinedCountTable.pl (best to restrict to some 100 top clones). ***Dependencies:*** R version 4.0.3 (rbioc_3-12), ggplot2 R packagePerl scripts equivalent to the corresponding bulkTCR script set - set up to work with cell ranger data. Script to extract only TRB information from scTCRseq data.