Skip to content

Tutorials Overview

This section provides comprehensive tutorials for using renalprog to analyze kidney cancer progression. Each tutorial is designed to be self-contained yet builds upon previous steps.

Tutorial Structure

🚀 Quick Start

New to renalprog? Start here!

📋 Complete Pipeline

Step-by-step walkthrough of the entire analysis pipeline:

  1. Data Processing: Download, filter, and preprocess TCGA data
  2. VAE Training: Train variational autoencoders
  3. Reconstruction Validation: Assess model quality
  4. Trajectory Generation: Create synthetic progression paths
  5. Classification: Stage prediction and biomarker discovery
  6. Enrichment Analysis: Pathway analysis with GSEA

🎨 Visualization

Learning Paths

For Biologists

If you're primarily interested in biological insights:

  1. Start with Quick Start
  2. Read Data Requirements to understand the data
  3. Follow Complete Pipeline end-to-end
  4. Focus on Enrichment Analysis for pathway interpretation
  5. Use Visualization Guide for publication figures

For Computational Scientists

If you want to customize models or develop new methods:

  1. Complete Quick Start
  2. Study VAE Training in detail
  3. Explore API Reference for implementation details
  4. Read Custom Models for extending functionality
  5. Review Architecture for design principles

For Reproducing Published Results

If you want to reproduce the paper:

Option 1: Using Pretrained Models (Recommended)

  1. Follow Using Pretrained Models tutorial
  2. This is the fastest and most accurate way to reproduce results
  3. Uses the exact models from the paper

Option 2: Training from Scratch

  1. Follow System Requirements
  2. Complete Data Preparation
  3. Execute Running the Pipeline
  4. Validate using Expected Results

Tutorial Conventions

Code Blocks

Python code to execute: 

from renalprog import dataset
data = dataset.load_data('path/to/data.csv')

Shell commands: 

python scripts/pipeline_steps/1_data_processing.py

Callouts

Note

Informational notes provide additional context.

Tip

Tips offer helpful suggestions and best practices.

Warning

Warnings highlight potential issues or common pitfalls.

Danger

Critical warnings about data loss or major errors.

Example

Example outputs or usage patterns.

File Paths

All file paths are relative to the repository root unless otherwise specified:

renalprog/
├── data/
│   ├── raw/           # Downloaded TCGA data
│   ├── interim/       # Intermediate processing outputs
│   └── processed/     # Final processed data
├── models/            # Trained models
├── reports/           # Analysis results
└── scripts/           # Pipeline scripts

Prerequisites

Before starting these tutorials, ensure you have:

  1. ✅ Installed renalprog (Installation Guide)
  2. ✅ Python 3.9+ and R 4.0+ available
  3. ✅ At least 16 GB RAM (8 GB minimum)
  4. ✅ 50+ GB free disk space
  5. ✅ (Optional) CUDA-capable GPU for faster training

Time Estimates

Tutorial Reading Time Execution Time Difficulty
Quick Start 10 min 15 min ⭐ Easy
Data Processing 15 min 30 min ⭐ Easy
VAE Training 20 min 2-4 hours* ⭐⭐ Moderate
Reconstruction 10 min 10 min ⭐ Easy
Trajectories 15 min 30 min ⭐⭐ Moderate
Classification 20 min 1 hour ⭐⭐ Moderate
Enrichment 25 min 2-6 hours* ⭐⭐⭐ Advanced
Visualization 15 min 30 min ⭐⭐ Moderate

* With GPU acceleration and 8+ CPU cores. Times may be significantly longer on older hardware.

Getting Help

If you encounter issues while following these tutorials:

  1. Check the Troubleshooting Guide
  2. Review the API Reference for function details
  3. Search GitHub Issues
  4. Ask in GitHub Discussions

Next Steps

Ready to begin? Start with the Quick Start Tutorial!