Comparative analysis of eccDNA and circRNA tools shows increased accuracy of tool combination

This repository contains the code and notebooks used in the study "Comparative analysis of eccDNA and circRNA tools shows increased accuracy of tool combination".

📁 Repository Structure

• benchmarking.ipynb: Main notebook performing the analysis and generating plots.
• functions/: Python scripts with modular functions for:

📥 How to Use This Repository

1. Clone the Repository

git clone https://github.com/ZabalaAitor/benchmarking
cd benchmarking

2. Download input data from Zenodo:

🔗 https://zenodo.org/records/15783792

After downloading, extract the contents and organize them as follows:

• data/ → should contain all BED files with raw and filtered outputs from circRNA and eccDNA detection tools.

• genomic_elements/ → must contain the GTF files and other genome annotation data required for genomic annotation analysis.

• mappability/ → should contain mappability values from UCSC 50-mer multi-track bigWig files (https://hgdownload.soe.ucsc.edu/gbdb/hs1/hoffmanMappability/)

• repeatmasker/ → must contain the txt file required for repeat element annotation analysis.

These four folders (data/, genomic_elements/, mappability/ and repeatmasker/) are essential for running the pipeline.

📊 Outputs

The ouputs of the benchmarking analysis are available at:
🔗 https://zenodo.org/records/15783794

Cite

Zabala, A., Ascensión, A. M., Prada-Luengo, I., & Otaegui, D. (2025). Comparative analysis of eccDNA and circRNA tools shows increased accuracy of tool combination. bioRxiv, 2025-07.