The Loading Lab 🧪

Welcome! This repo is a collection of lightweight, sandbox-style demos for common life-science data types.
Each folder is self-contained and shows how to:

1. Load the data into Python
2. Do 1–3 basic EDA plots or summaries
3. Point you to next steps if you want to dive deeper

The goal: lower the barrier to entry. Once you can load and peek, you can explore on your own.

📌 Notes

• Note the Disclaimer!
  • The purpose of these demo notebooks are just to get you up to speed and running some fun analyses on common lab data outputs.
  • Don't stop there if (when!) you reach the point of putting this into production.
• Note the Roadmap!
  • I'm completing demos in descending order of the historical level of fun I've had working with the data (...because, why not?)
  • If there's something particular you'd like me to prioritize, please let me know! Always happy to have an extra data point to help prioritize.
• Environments
  • I'm making 1 per data type to avoid requirement conflicts.
• Data
  • Links the the README of the respective folders.
  • I'll make an extra "data download" notebook as/when convenient.

⚠️ Disclaimer

The goal of The Loading Lab is to help you get started — fast. These are basic, stand-alone notebooks meant to lower the barrier to entry and make the first steps with real data fun and approachable.

• The notebooks are deliberately written in a quick, exploratory style — not how I’d structure production-quality pipelines in real work.

• Once you’re comfortable loading and plotting data here, the next step is to adopt best practices for serious projects, such as:

  • Using the command line and scripts for reproducibility

  • Writing modular functions and packages instead of monolithic notebooks

  • Adding version control (Git) for your analysis code

  • Managing environments with conda/mamba or pip and pinning versions

  • Building workflows with Snakemake / Nextflow / CWL for maintainability

  • Using Docker/containers when you need portability across systems

  • Writing tests and using continuous integration for long-term reliability

I like the notebook approach here because it makes the repo easy to read and follow along without setup overhead — you can see the logic and outputs inline.

So treat these as “sandbox demos”: once you’re ready to do your own research or production work, start layering on the right habits and tools for scalability and rigor.

Enjoy, and I hope this helps you learn and have fun!

✅ Modules included

• cdisc_adam/ — ADaM .xpt & define.xml files
• cdisc_stdm/ — STDM .xpt & define.xml files
• flow_cytometry/ — FCS files (CyTOF/flow): channels, FSC/SSC, histograms

🏗️ Road Map / Under Construction

(Just thinking out loud here....)

• rnaseq_bulk/ — bulk RNA-seq: counts → QC → toy DE
• rnaseq_single_cell/ — scRNA-seq: AnnData, PCA/UMAP, quick clustering
• proteomics_ms/ — proteomics matrices: PCA, heatmap
• microscopy/ — load microscopy TIFF/PNG, display, histogram, simple segmentation
• vcf_variants/ — parse VCF, count SNPs/indels, variant density plots
• elisa_plate/ — plate reader CSV → heatmap
• Some QA/QC stuff would be fun --> maybe put it in the specific folder for that data type? Or give QA/QC work it's own folder.
• Automations - same question as the QA/QC.