A Comprehensive Benchmark for RNA 3D Structure-Function Modeling

2025-03-27

AI Summary

RNA structure–function relationship modeling has long been hindered by the absence of standardized, accessible benchmark datasets. To address this, we introduce RNA3D-Bench—the first open-source benchmark suite specifically designed for RNA 3D structure–function tasks—comprising seven modular, extensible, and reproducible datasets covering diverse RNA functional prediction tasks. We propose the first unified evaluation framework supporting end-to-end workflows: data encoding, dataset splitting, model training, and multi-metric evaluation—fully compatible with the PyTorch ecosystem and standard data interfaces. Built upon rnaglib, the suite integrates graph neural networks (GNNs) as baseline models and provides systematic performance baselines across all tasks. The project is community-oriented, enabling contributions and customization; its codebase, documentation, and pre-trained models are fully open-sourced. RNA3D-Bench has become a foundational resource widely adopted in AI-driven RNA structural biology research.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: 3D Computer Vision
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

The RNA structure-function relationship has recently garnered significant attention within the deep learning community, promising to grow in importance as nucleic acid 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 set of seven benchmarking datasets for RNA structure-function prediction, designed to address this gap. Our library builds on the established Python library rnaglib, and offers easy data distribution and encoding, splitters and evaluation methods, providing a convenient all-in-one framework for comparing models. Datasets are implemented in a fully modular and reproducible manner, facilitating for community contributions and customization. Finally, we provide initial baseline results for all tasks using a graph neural network. Source code: https://github.com/cgoliver/rnaglib Documentation: https://rnaglib.org

Problem

• Lack of standardized benchmarks for RNA 3D structure deep learning
• Need for accessible tools to predict RNA structure-function relationships
• Absence of modular datasets for RNA model comparison

Innovation

• Introduces seven RNA structure-function benchmarking datasets
• Builds on rnaglib for modular data handling
• Uses graph neural network for baseline results