11  Data Analysis Track

Welcome to the Data Analysis demonstration track. Building on the MCP concepts from the Paper Writing track, here you’ll apply Claude Code to more complex scientific computing tasks.

11.1 What We’ll Learn

This track demonstrates how Claude Code can help with:

  • MCPs (Model Context Protocol): Extending Claude Code with tools for literature search, databases, and more
  • Data acquisition: Downloading from databases and APIs
  • Data processing: Cleaning, transforming, restructuring
  • Analysis: Statistical tests, pattern finding, summarization
  • Visualization: Creating publication-quality figures
  • Automation: Building reproducible pipelines
Introducing MCPs

A key goal of this track is to introduce MCPs—plugins that give Claude Code direct access to external services like PubMed, databases, and specialized tools. Instead of writing API code yourself, you simply ask Claude what you want, and the MCP handles the rest.

11.2 Choose Your Demo

We’ve prepared three demonstrations. You can work through one or all:

11.2.1 Demo 1: Meta-Analysis of Cell Type Mapping Tools

Goal: Determine which cell type mapping tools are most used in practice by mining scientific literature.

You’ll learn: - Searching academic literature programmatically - Working with MCPs (Model Context Protocol) for literature search - Text mining and parsing - Creating summary visualizations

Best for: Those interested in literature mining, bibliometrics, or single-cell biology

Go to Meta-Analysis Demo

11.2.2 Demo 2: Protein Structure Visualization with PyMol

Goal: Create publication-quality visualizations of a protein binding site without prior PyMol knowledge.

You’ll learn: - Using Claude Code to learn new tools - Structural biology basics - PyMol scripting - Generating figures

Best for: Those interested in structural biology or learning new software tools

Go to PyMol Demo

11.2.3 Demo 3: Reverse-Engineering Oligo Design (TRISCO)

Goal: Understand and reproduce the oligo design logic from the Kanatani et al. TRISCO method.

You’ll learn: - Reading and interpreting methods sections - Primer and oligo design principles - Reverse-engineering experimental protocols - Validating designs computationally

Best for: Those interested in molecular biology methods or experimental design

Go to TRISCO Demo

11.3 Prerequisites

For these demos, you’ll need:

  • Claude Code installed and working
  • Python 3.8+ with pip
  • Internet connection (for data downloads)
  • Basic understanding of your chosen domain (helpful but not required)

Some demos may have additional requirements noted at the start.

11.4 Let’s Begin

Choose a demo above, or start with the first one: Meta-Analysis of Cell Type Mapping Tools.