creating-r-research-projects_skill

Overview

This skill sets up reproducible R research projects and automates analysis, package management, and report generation. It creates a standardized project layout, configures a reproducible environment, and produces publication-ready figures and documents. The focus is on scripted, repeatable workflows for statistics and bioinformatics.

How this skill works

The skill scaffolds an R project directory (data/, scripts/, results/, reports/) and initializes renv or an RStudio project file for environment reproducibility. It installs required CRAN and Bioconductor packages, generates analysis scripts for data import, cleaning, statistical modeling, and visualization, and saves outputs (CSVs, plots, model summaries). Finally it creates an R Markdown or Quarto report and renders it to HTML or PDF with embedded methods, figures, and results.

When to use it

• You need a reproducible R project scaffold for a new analysis
• Analyze CSVs, gene expression, or omics datasets with R
• Run statistical tests, regression models, or differential expression workflows
• Produce publication-ready plots and a methods/results report
• Manage and lock package dependencies for long-term reproducibility

Best practices

• Use renv to capture and restore package versions for reproducibility
• Keep all raw data in data/ and write scripts that create processed files programmatically
• Script the complete analysis end-to-end; avoid manual interactive steps unless requested
• Save all outputs (plots, tables, model objects) to results/ with descriptive filenames
• Write a single R Markdown or Quarto document that knits figures and tables automatically

Example use cases

• Create a new R project to analyze a clinical CSV and produce regression output and plots
• Run differential expression using DESeq2 and output PCA and volcano plots with an HTML report
• Set up a microbiome analysis pipeline with tidyverse, phyloseq, and save ordination plots
• Generate a reproducible report that documents methods, code, and results for peer review
• Lock dependencies with renv and provide a restore-ready environment for collaborators