Skip to content

Workshop Wrap-up

Congratulations! You have successfully built a complete, reproducible RNA-seq pipeline using Snakemake, starting from raw Saccharomyces cerevisiae FASTQ files and finishing with a multi-sample count matrix and an interactive MultiQC quality report.

Transitioning from ad-hoc shell scripts to a rule-based workflow manager is one of the most impactful changes you can make as a computational researcher. What began as a single FastQC rule is now a fully parallelised, cluster-ready, version-locked pipeline that handles any number of samples by changing one list.

Cleaning Up

Once you are done exploring, you can free up disk space by removing the workshop data. Choose the level of cleanup that suits you:

# Remove all pipeline outputs (safe to regenerate any time)
rm -rf results/

# Remove downloaded FASTQ files (~1.5 GB)
rm -rf data/

# Remove the genome and STAR index (~170 MB)
rm -rf genome/

# Remove cached Conda environments and the Apptainer .sif image
rm -rf .snakemake/conda/
rm -rf .snakemake/singularity/

# Or remove the entire project directory at once
rm -rf ~/snakemake_workshop

Double-check before deleting

rm -rf is permanent. Make sure you are in (or referencing) the right directory before running it.

What You Have Achieved

Over the course of this workshop, you have learned how to:

  • Write Snakemake Rules: Define pipeline steps with input:, output:, and shell: blocks, and understand how Snakemake traces dependencies backward from a target file to determine exactly what needs to run.
  • Generalise with Wildcards: Replace hardcoded filenames with {sample} wildcards and use expand() to apply a single rule across all samples automatically, so adding a new sample means changing one line.
  • Manage Computational Resources: Declare threads: and resources: per rule so Snakemake parallelises jobs intelligently without overloading your machine.
  • Scale to HPC: Submit the exact same Snakefile to a SLURM cluster using snakemake-executor-plugin-slurm, with threads: and mem_mb automatically translated into #SBATCH directives.
  • Ensure Reproducibility: Pin software environments per rule with conda: directives and version-locked YAML files, so your pipeline produces the same results on any machine, today, next year, and when a reviewer asks you to revisit your analysis.

Why This Matters

The scientific computing landscape is shifting. Journals and granting agencies increasingly mandate strict computational reproducibility. By expressing your analysis as a Snakemake workflow, you are no longer sharing a script that might work on someone else's machine. You are sharing a complete, self-documenting pipeline that:

  • Skips steps whose outputs are already up to date.
  • Parallelises automatically across whatever hardware is available.
  • Installs the exact tools and versions used to generate your published results.
  • Scales from a laptop to a 1000-core HPC cluster by changing a single flag.

Next Steps & Resources

As you begin applying Snakemake to your own research, keep these resources bookmarked:

  • Snakemake Official Documentation — The definitive reference for rule syntax, directives, configuration files, and advanced features like checkpoints and report generation.
  • Snakemake Wrappers — A community library of ready-to-use wrappers for hundreds of bioinformatics tools (STAR, fastp, featureCounts, and many more). Use a wrapper and the conda: environment is already written for you.
  • Snakemake Workflows — A curated collection of peer-reviewed, production-grade Snakemake pipelines for RNA-seq, ChIP-seq, variant calling, and more. A great starting point for real analyses.

Thank you for taking the time to learn Snakemake with us. We look forward to seeing the reproducible, scalable science you publish next!

— Hans and the Nevada Bioinformatics Center Team