Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RDesign: Hierarchical Data-efficient Representation Learning for Tertiary Structure-based RNA Design

Cheng Tan, Yijie Zhang, Zhangyang Gao, Bozhen Hu, Siyuan Li, Zicheng Liu, Stan Z. Li

TL;DR

This study crafted a large, well-curated benchmark dataset and designed a comprehensive structural modeling approach to represent the complex RNA tertiary structure and proposed a hierarchical data-efficient representation learning framework that learns structural representations through contrastive learning at both cluster-level and sample-level to fully leverage the limited data.

Abstract

While artificial intelligence has made remarkable strides in revealing the relationship between biological macromolecules' primary sequence and tertiary structure, designing RNA sequences based on specified tertiary structures remains challenging. Though existing approaches in protein design have thoroughly explored structure-to-sequence dependencies in proteins, RNA design still confronts difficulties due to structural complexity and data scarcity. Moreover, direct transplantation of protein design methodologies into RNA design fails to achieve satisfactory outcomes although sharing similar structural components. In this study, we aim to systematically construct a data-driven RNA design pipeline. We crafted a large, well-curated benchmark dataset and designed a comprehensive structural modeling approach to represent the complex RNA tertiary structure. More importantly, we proposed a hierarchical data-efficient representation learning framework that learns structural representations through contrastive learning at both cluster-level and sample-level to fully leverage the limited data. By constraining data representations within a limited hyperspherical space, the intrinsic relationships between data points could be explicitly imposed. Moreover, we incorporated extracted secondary structures with base pairs as prior knowledge to facilitate the RNA design process. Extensive experiments demonstrate the effectiveness of our proposed method, providing a reliable baseline for future RNA design tasks. The source code and benchmark dataset are available at https://github.com/A4Bio/RDesign.

RDesign: Hierarchical Data-efficient Representation Learning for Tertiary Structure-based RNA Design

TL;DR

This study crafted a large, well-curated benchmark dataset and designed a comprehensive structural modeling approach to represent the complex RNA tertiary structure and proposed a hierarchical data-efficient representation learning framework that learns structural representations through contrastive learning at both cluster-level and sample-level to fully leverage the limited data.

Abstract

While artificial intelligence has made remarkable strides in revealing the relationship between biological macromolecules' primary sequence and tertiary structure, designing RNA sequences based on specified tertiary structures remains challenging. Though existing approaches in protein design have thoroughly explored structure-to-sequence dependencies in proteins, RNA design still confronts difficulties due to structural complexity and data scarcity. Moreover, direct transplantation of protein design methodologies into RNA design fails to achieve satisfactory outcomes although sharing similar structural components. In this study, we aim to systematically construct a data-driven RNA design pipeline. We crafted a large, well-curated benchmark dataset and designed a comprehensive structural modeling approach to represent the complex RNA tertiary structure. More importantly, we proposed a hierarchical data-efficient representation learning framework that learns structural representations through contrastive learning at both cluster-level and sample-level to fully leverage the limited data. By constraining data representations within a limited hyperspherical space, the intrinsic relationships between data points could be explicitly imposed. Moreover, we incorporated extracted secondary structures with base pairs as prior knowledge to facilitate the RNA design process. Extensive experiments demonstrate the effectiveness of our proposed method, providing a reliable baseline for future RNA design tasks. The source code and benchmark dataset are available at https://github.com/A4Bio/RDesign.
Paper Structure (39 sections, 15 equations, 28 figures, 16 tables)

This paper contains 39 sections, 15 equations, 28 figures, 16 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Biomolecular Engineering
  4. Protein Design
  5. RNA Design
  6. Methods
  7. Preliminaries
  8. Comprehensive RNA Tertiary Structure Modeling
  9. Intra-nucleotide level
  10. Inter-nucleotide level
  11. Hierarchical Data-efficient Representation Learning
  12. Secondary Structure Imposing Strategy
  13. Experiments
  14. Standard Tertiary Structure-based RNA Design
  15. Evaluate the Generalization on Rfam and RNA-Puzzles
  16. ...and 24 more sections

Figures (28)

  • Figure 1: The schematic diagrams of RNA primary, secondary and tertiary structures.
  • Figure 2: Brief view of RNA sequence and secondary structure.
  • Figure 3: An example of RNA structure.
  • Figure 4: The different levels of hierarchical representation learning. Green arrows denote positive pairs tend to attract each other, and the black arrow denotes negative pairs tend to repel each other.
  • Figure 5: Visualization of RDesign's designed examples.
  • ...and 23 more figures