1  Introduction

1.1 What We’re Going to Do

In Part 1, you’re going to install a few tools and then run a real single-cell RNA-seq analysis, including loading a count matrix, filtering and normalizing the data, clustering cells, and generating a UMAP visualization. You’ll do all of this interactively in Positron, an IDE we use to organize coding projects, write and run code, and interact with AI assistants. At the end you will have set up a basic project, created and run some preliminary analysis, and generated a Quarto document that doubles as both script and shareable report.

Along the way, you’ll learn to use the essential features of Positron that are used every day: the file explorer for navigating your project, the console for running code interactively, the environment pane for inspecting your data, the plots pane for viewing figures, and the terminal for running commands.

By the end, you’ll have a working project with a rendered HTML report that you could share with anyone in the lab. That’s the starting point for every analysis you’ll do here.

1.2 How We Work in the Lab

Most analysis in the Musser Lab is written in Quarto documents (.qmd files). A Quarto document is both a script, containing code chunks you can run interactively, and a report, which includes annotations that help readers (and yourself!) remember and understand what the code is doing. In addition to running code interactively, you can also render the Quarto document to produce a clean HTML (or PDF) document with your results, figures, and narrative included. This means your analysis and your documentation are always in sync, and there’s no separate step of copying figures into a report or updating numbers by hand.

Our projects typically use both R and Python, though usually one language per script. R is used for statistical analysis, plotting, and the many well-established bioinformatics packages (Seurat, DESeq2, limma). Python was classically used for manipulating text files, working with sequence data, and running pipelines that manipulated files in bash. Increasingly, many modern single-cell and biological data science tools are being developed in Python. R remains essential for plenty of analyses, so lab members will need to be competent in both. Fortunately, Positron handles both languages equally well, so you can switch between them as needed.

1.3 The Tools

Here’s a quick overview of the tools you’ll encounter in this guide and how they fit together. You don’t need to understand all of this now — you’ll learn each tool as you need it.

┌───────────────────────────────────────────────────┐
│                      Positron                     │
│          (where you write and run code)           │
│                                                   │
│   ┌───────────┐  ┌───────────┐  ┌─────────────┐   │
│   │   Conda   │  │   renv    │  │ Claude Code │   │
│   │  (Python  │  │    (R     │  │ (AI coding  │   │
│   │ packages) │  │ packages) │  │  assistant) │   │
│   └───────────┘  └───────────┘  └─────────────┘   │
│                                                   │
│   ┌───────────────────────────────────────────┐   │
│   │               Git / GitHub                │   │
│   │   (version control and collaboration)     │   │
│   └───────────────────────────────────────────┘   │
└───────────────────────────────────────────────────┘

Positron is your home base — the application where you write code, explore data, view plots, and manage your project files. It’s built by the makers of RStudio but designed for both R and Python.

Conda and renv are environment managers. They keep track of which packages (and which versions) each project uses, so your analysis works the same way on any computer. Conda handles Python; renv handles R. You’ll set up renv in your first project and learn conda later in Part 2.

Git and GitHub let you track changes to your code over time and share it with collaborators. Think of Git as an unlimited undo history for your project, and GitHub as the place where that history lives online. You’ll learn these in Part 2.

Claude Code is an AI coding assistant that integrates into Positron. It can help you write code, debug errors, explain unfamiliar functions, and plan analyses. You’ll install it in the next chapter and see it in action throughout the guide.

1.4 What You’ll Learn in Part 1

After working through the next two chapters, you will have:

  • Installed Positron, R, Quarto, and Claude Code
  • Created a project with a proper folder structure
  • Run a complete single-cell analysis interactively in Positron, learning the IDE along the way
  • Rendered your analysis into a self-contained HTML report
  • Used renv to manage your R packages

That’s enough to start working on real lab projects. Parts 2–4 will deepen your understanding of each tool and introduce version control, collaboration workflows, and more advanced techniques — but you can come back to those as you need them.

Let’s start by installing the tools.