README.md

---
license: mit
task_categories:
- graph-ml
tags:
- biology
- 3D
---

# Dataset Card for Dataset Name

<!-- Provide a quick summary of the dataset. -->

This dataset contains the tasks provided by _rnaglib_, a benchmarking suite for RNA structure-function modelling.

## Dataset Details

### Dataset Description

<!-- Provide a longer summary of what this dataset is. -->
This dataset is a Python package wrapping RNA benchmark datasets and tasks. Data access, preprocessing, and task-specific pipelines are implemented in code and not fully expressed through this metadata schema. See documentation at https://github.com/cgoliver/rnaglib.


- **Curated by:** [Luis Wyss, Vincent Mallet, Wissam Karroucha, Carlos Oliver, Karsten Borgwardt]
- **License:** [MIT license]

### Dataset Sources [optional]

<!-- Provide the basic links for the dataset. -->

- **Repository:** https://github.com/cgoliver/rnaglib
- **Preprint:** https://arxiv.org/pdf/2503.21681
- **Documentation:** https://rnaglib.org/en/latest/

## Uses

The intended use of these dataset is to train models to elucidate RNA structure-function relationships.

## Dataset Structure

<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->

Detailed information can be found in the linked paper. This repo contains nine tasks, each coming with their own splitting. The splitting is either based on published splits or on structural dissimilarity.


## Dataset Creation

### Curation Rationale

<!-- Motivation for the creation of this dataset. -->

We provide datasets replicating published research or generate our own datasets ensuring effective QC, filtering, redundancy removal and splitting.

### Source Data

<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->

All data originates from RNA structures pulled from the PDB.

#### Data Collection and Processing

<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->

Details can be found in the linked paper.

#### Who are the source data producers?

<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->

The data originates from the PDB, the data producers are the contributors to the PDB.

### Annotations [optional]

The data is saved as graphs and annotated with features extracted from th PDB.



#### Who are the annotators?

<!-- This section describes the people or systems who created the annotations. -->

_rnaglib_ is used to process the data.

#### Personal and Sensitive Information

<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->

N/A

## Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

N/A


## Citation

<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

@article{wyss2025comprehensivebenchmarkrna3d,
      title={A Comprehensive Benchmark for RNA 3D Structure-Function Modeling}, 
      author={Luis Wyss and Vincent Mallet and Wissam Karroucha and Karsten Borgwardt and Carlos Oliver},
      year={2025},
      eprint={2503.21681},
      archivePrefix={arXiv},
      primaryClass={q-bio.BM},
      url={https://arxiv.org/abs/2503.21681}, 
}



## Dataset Card Contact

wyss@biochem.mpg.de