Standardized Snakemake Workflow for Survival Analysis

This repository provides a Snakemake-based workflow for survival analysis using datasets from SurvSet and external CSV files. The workflow automates data preprocessing, model training, cross-validation, and evaluation using survival models from SurvHive.

🚀 Overview

This workflow follows Snakemake best practices and ensures reproducibility across different computational environments. It allows users to:

• Load survival datasets (SurvSet or external sources)
• Preprocess and split the data
• Perform cross-validation with multiple survival models
• Evaluate model performance using the Concordance Index
• Generate a final report summarizing results

🛠️ Installation & Dependencies

To run this workflow, you need:

• Snakemake (v7.0+)
• Conda (to manage dependencies)

Install Snakemake & Conda

If not installed, set up Miniconda and Snakemake:

conda install -c conda-forge mamba  # Faster Conda environment management
mamba create -n snakemake_env -c conda-forge snakemake
conda activate snakemake_env

📂 Repository Structure

├── config/                 # Configuration files
│   ├── config.yaml.example # Example configuration file
│   ├── README.md           # Configuration guide
├── data/                   # Input datasets (ignored in Git)
├── logs/                   # Log files (ignored in Git)
├── results/                # Processed outputs
├── scripts/                # Python scripts for preprocessing & training
├── workflow/               # Snakemake rules and pipeline logic
│   ├── Snakefile           # Main Snakemake workflow
│   ├── rules/              # Individual Snakemake rule definitions
│   ├── envs/               # Conda environments (optional)
├── .snakemake/             # Snakemake cache (ignored in Git)
├── .snakemake-workflow-catalog.yml  # Snakemake catalog metadata
├── .gitignore              # Ignore unnecessary files
└── README.md               # This document

⚙️ Configuring the Workflow

Before running, create a custom configuration file:

cp config/config.yaml.example config/config.yaml

Edit config/config.yaml to specify:

• Datasets (SurvSet or external CSVs)
• Models to use for training
• Dataset structure (Required columns: pid, event, time)

▶️ Running the Workflow

Once configured, execute Snakemake:

snakemake --use-conda --cores <n>

• <n>: Number of CPU cores (e.g., --cores 4)
• Use --configfile config/config.yaml to specify a custom config file

⚠️ First-Time Setup Warning

During the first run, Snakemake will create Conda environments for different survival models, which may take some time due to the number of required dependencies. Once built, subsequent runs will be much faster.

Running a Specific Rule

snakemake --use-conda preprocess_and_split

📖 References & Citations

This workflow is built upon the following frameworks and datasets:

• SurvHive: A package for survival model optimization and evaluation.
  Citation: Birolo, Giovanni, et al. "SurvHive: a package to consistently access multiple survival-analysis packages." arXiv preprint arXiv:2502.02223 (2025).
  GitHub: SurvHive Repository

• SurvSet: An open-source time-to-event dataset repository.
  Citation: Drysdale, Erik. "SurvSet: An open-source time-to-event dataset repository." arXiv preprint arXiv:2203.03094 (2022).
  GitHub: SurvSet Repository

---

For additional support, refer to the config README (config/README.md) or contact the workflow maintainer. 🚀