These scripts were used to pre-process the raw seqencing data, and analyse the resultant count matrix, as well as to prepare figures in the manuscript: [Astrocytes in the mouse suprachiasmatic nuclei respond directly to glucocorticoids feedback](bioRxiv link)
Here is a summary of all the analysis steps.
The scripts for each step are contained within the respective directories along with an associated readme file, which contain individual installation and usage requirements, whenever necesary. Down below are some global installation instructions and system requirements.
-
pre_processing 10X Cellranger-based scripts for obtaining single-cell gene-level count matrixes from fastq files.
-
analysis_pipeline Contains generic NextFlow-based pipeline for basic sn-RNAseq data analysis that includes QC reporting, QC filtering including doublet detection, clustering, and marker-gene identification, merging/integration of multiple samples, trajectory analysis, ssGSEA analysis, and ligand-receptor analysis. These are implemented using Seurat (v4) and monocle3, escape (v1.8.0), in addition to other R-packages.
Most of these scripts were written to work in a HPC running Debian GNU/Linux 11 (bullseye). Most scripts are wrapped within an additional Slurm job-submission script (sbatch), which can be bypassed if running on a local machine. Moreover, all the paths needs to be adjusted as required.
All the packages were installed within conda environments. Instructions/commands to generate these conda environments are located in installation/*.yaml
Files included are:
-
Downsampled version of the two Seurat objects corresponding to two time points in our dataset that can be used as inputs to the scripts in the analysis_pipeline
-
Downsampled version of the Morris et. al. dataset that contains our annotations. This can be used as input for ssGSEA ssGSEA and for ligand-receptor analyses.