A Comprehensive Library for RNA Structure-Function Modeling

Luis Wyss, Vincent Mallet, Wissam Karroucha, Karsten Borgwardt, Carlos Oliver

Track: long paper (up to 6 pages)

Keywords: RNA, structure, learning, benchmark

TL;DR: We propose several tasks relevant to the functional annotation of RNA 3D structure, and a principled and easy way to load and evaluate them for machine learning models benchmarking.

Abstract:

The RNA structure-function relationship has recently garnered significant attention within the deep learning community, promising to grow in importance as nucleotide structure models advance.
However, the absence of standardized and accessible benchmarks for deep learning on RNA 3D structures has impeded the development of models for RNA functional characteristics.

In this work, we introduce a comprehensive set of benchmarking datasets for RNA structure modeling, designed to address this gap.
Our library includes easy data distribution and encoding, splitters and evaluation methods, providing a robust suite for comparing models.
Beyond the proposed tasks, our library is modular and thereby can easily be tailored by researchers to their question at hand. We provide experiments highlighting the ease of use of our library.

Submission Number: 7