Mouse CNS single-cell RNA-seq analysis with Seurat

This repository contains a cleaned and reproducible Seurat workflow for the PanglaoDB dataset SRA667466_SRS3059954. The analysis covers quality control, MAD-based cell filtering, log normalization, optional cell-cycle scoring, highly variable gene selection, PCA, comparison of clustering solutions, marker discovery, and marker-supported cell-type annotation.

• Replaced absolute Windows paths with relative project paths.
• Moved package installation outside the analysis script.
• Removed duplicated cell-cycle, PCA, clustering, marker, and annotation blocks.
• Fixed inconsistent clustering dimensions and an accidental use of the wrong object in the 30-PC solution.
• Replaced repeated code with reusable functions.
• Added systematic output names and automatic saving of plots, tables, the final Seurat object, and sessionInfo().
• Preserved the complete original script in legacy/.

.
├── analysis/run_analysis.R       # Complete  pipeline
├── R/functions.R                 # Reusable analysis functions
├── config/config.R               # Parameters and clustering solutions
├── data/raw/                     # Local input files 
├── .gitignore
└── README.md

Place this file in data/raw/:

SRA667466_SRS3059954.sparse.RData

Cell-cycle scoring is optional. To enable it, also place:

cc.genes.mouse.RDS

The cell-cycle file must contain two named elements: s.genes and g2m.genes.

Open R or a terminal in the repository root and run:

source("install_packages.R")
source("analysis/run_analysis.R")

or:

Rscript install_packages.R
Rscript analysis/run_analysis.R

The default configuration evaluates:

• 10 PCs, resolution 0.5
• 20 PCs, resolution 0.5
• 20 PCs, resolution 0.8
• 25 PCs, resolution 0.5
• 30 PCs, resolution 0.5

The final solution is set to 25 PCs and resolution 0.5, matching the selection in the original project. Change final_solution in config/config.R to select another solution.

The generated annotations are based on overlap between cluster markers and a curated mouse CNS marker panel. They are intended as annotation support and must be reviewed using marker specificity, expression plots, biological context, and potential technical biases.

Large source data files are intentionally excluded from Git. See data/README.md for placement instructions.