Transcriptional dynamics during juvenile and adult spermatogenesis

This repository contains scripts for data processing, analysis and figure generation using scRNA-Seq, bulk RNAseq and CUT&RUN data.

How to work with the data

The analyses scripts load in a SingleCellExperiment object that contains either the cell ranger filtered cells or the EmptyDrops filtered cells (see Methods section of the publication). These single-cell RNA sequencing data have been deposited on ArrayExpress under the accession number E-MTAB-6946. The following code chunk downloads the data and stores them in a SingleCellExperiment object that can be used to reproduce the analysis performed for this project.

1. Obtaining the data

In the first step, clone this repository to your local Github folder:

cd ~/Github
git clone https://github.com/MarioniLab/Spermatogenesis2018.git

To obtain the cell ranger filtered data use:

setwd("~/GitHub/Spermatogenesis2018/")

# Raw cellranger filtered transcriptomes
download.file("https://www.ebi.ac.uk/arrayexpress/files/E-MTAB-6946/E-MTAB-6946.processed.1.zip", 
                destfile = "cellranger_raw.zip")
unzip("cellranger_raw.zip")
file.remove("cellranger_raw.zip")

# Cellranger filtered metadata
download.file("https://www.ebi.ac.uk/arrayexpress/files/E-MTAB-6946/E-MTAB-6946.processed.3.zip", 
                destfile = "cellranger_metadata.zip")
unzip("cellranger_metadata.zip") 
file.remove("cellranger_metadata.zip")

# Cellranger filtered gene names
download.file("https://www.ebi.ac.uk/arrayexpress/files/E-MTAB-6946/E-MTAB-6946.processed.2.zip", 
                destfile = "cellranger_genes.zip")
unzip("cellranger_genes.zip") 
file.remove("cellranger_genes.zip")

To obtain the EmptyDrops filtered data use:

# Raw empty drops filtered transcriptomes
download.file("https://www.ebi.ac.uk/arrayexpress/files/E-MTAB-6946/E-MTAB-6946.processed.4.zip", 
                destfile = "EmptyDrops_raw.zip")
unzip("EmptyDrops_raw.zip")
file.remove("EmptyDrops_raw.zip")
               
# EmptyDrops filtered metadata
download.file("https://www.ebi.ac.uk/arrayexpress/files/E-MTAB-6946/E-MTAB-6946.processed.6.zip", 
                destfile = "EmptyDrops_metadata.zip")
unzip("EmptyDrops_metadata.zip") 
file.remove("EmptyDrops_metadata.zip")

# EmptyDrops filtered gene names
download.file("https://www.ebi.ac.uk/arrayexpress/files/E-MTAB-6946/E-MTAB-6946.processed.5.zip", 
                destfile = "EmptyDrops_genes.zip")
unzip("EmptyDrops_genes.zip") 
file.remove("EmptyDrops_genes.zip")

2. Creating the SingleCellExperiment object

In the next step, we create a SingleCellExperiment object by combining the raw counts with matched metadata.

# Create SingleCellExperiment object
library(SingleCellExperiment)
library(Matrix)

# Cellranger filtered data
cellranger_raw <- readMM("raw_counts.mtx")
cellranger_metadata <- read.table("cell_metadata.txt", sep = " ")
cellranger_genes <- read.table("genes.tsv", sep = "\t", header = TRUE)
sce_cellranger <- SingleCellExperiment(assays = list(counts = as(cellranger_raw,
                                                                  "dgCMatrix")), 
                                       colData = cellranger_metadata,
                                       rowData = cellranger_genes) 
rownames(sce_cellranger) <- rowData(sce_cellranger)$ID                                      
saveRDS(sce_cellranger, "SCE_all.rds")

# EmptyDrops filtered data
EmptyDrops_raw <- readMM("raw_counts_emptyDrops.mtx")
EmptyDrops_metadata <- read.table("cell_metadata_emptyDrops.txt", sep = " ")
EmptyDrops_genes <- read.table("genes_emptyDrops.tsv", sep = "\t", header = TRUE)
sce_EmptyDrops <- SingleCellExperiment(assays = list(counts = as(EmptyDrops_raw,
                                                                  "dgCMatrix")), 
                                       colData = EmptyDrops_metadata,
                                       rowData = EmptyDrops_genes) 
rownames(sce_EmptyDrops) <- rowData(sce_EmptyDrops)$ID 
saveRDS(sce_EmptyDrops, "SCE_emptyDrops.rds")

3. Further processing

Of note: the analysis performed here was done using packages of Bioconductor version 3.8 and R version 3.5. Some of the packages have changed since then and some of the functions are not longer supported or moved to other packages.

You can install packages of Bioconductor 3.8 running:

BiocManager::install(version = "3.8")

To further process the data (normalization, dimensionality reduction, ...), please follow either the Filtering (from line 166) or the EmptyDrops (from line 262) scripts.

Reading in the data

All scripts in this repository load the SingleCellExperiment objects from where I saved them. To load in the objects, please change the paths to setwd("~/GitHub/Spermatogenesis2018/") or the path where you saved the objects. They can be read in using sce <- readRDS("SCE_all.rds").

Analysis scripts

The following folders contain scripts for data processing and analysis. A short description can be found below:

• Preprocessing contains quality control and normalization scripts for scRNA-Seq and bulk RNA-Seq data, as well as a script to process the data from Chen et al., Cell Research, 2018.

• All scripts to reproduce the individual figures (main and supplements) can be found in Figures.

• Data contains a file listing mouse gene IDs, Symbols and corresponding chromosome information.

• Functions contains an R script with wrapper functions for simple tasks throughout the analysis. Furthermore, this folder contains a script to count and normalize reads of CUT&RUN and ChIP-Seq libraries in promoter regions.

• Analysis contains .Rmd files with additional analysis performed on the data. These analyses have not been used to present data in the main manuscript.

• Shiny contains scripts to prepare data for the shiny app visualization and the server.R and ui.R files.

• Shiny_ED: Same as Shiny but for emptyDrops data.

The paper can be found at:

Staged developmental mapping and X chromosome transcriptional dynamics during mouse spermatogenesis