Related Experiment Video
Updated: Jul 9, 2026
05:41
2D-HELS MS Seq: A General LC-MS-Based Method for Direct and de novo Sequencing of RNA Mixtures with Different Nucleotide Modifications
Published on: July 10, 2020
Identifying RNA ac4C Modification Sites via Pseudo-Nucleotide Fingerprint Encoding and Multi-Scale Feature
Yiming Wang1, Fan Mo2,3, Yun Sha1
1School of Information Engineering, Beijing Institute of Petrochemical Technology, Beijing, 102617, China.
Summary
This study introduces DFM-ac4C, a new computational tool for accurately identifying N4-acetylcytidine (ac4C) RNA modifications. This advancement aids in understanding RNA regulation and cancer progression, offering a more effective method for site prediction.
Area of Science:
- Molecular Biology
- Bioinformatics
- Computational Biology
Background:
- N4-acetylcytidine (ac4C) RNA modifications are crucial for gene regulation, mRNA stability, and stress response.
- Dysregulated ac4C modifications are linked to cancer development, indicating potential as biomarkers and therapeutic targets.
- Current computational methods for predicting RNA modification sites struggle with multiscale nucleotide interactions and long-range dependencies.
Purpose of the Study:
- To develop a novel computational framework, DFM-ac4C, for the accurate prediction of ac4C modification sites in RNA.
- To overcome the limitations of existing methods by integrating diverse features and advanced deep learning architectures.
Main Methods:
- DFM-ac4C integrates pseudo-nucleotide fingerprint features (Kidera factors) with global contextual embeddings (Evo2, DNAshape).
- An attention-based fusion architecture dynamically combines local sequence information, fingerprint descriptors, and global embeddings.
- The framework is designed to capture complex, multiscale characteristics of nucleotide interaction sites.
Main Results:
- DFM-ac4C significantly outperforms existing models in predicting ac4C modification sites.
- Achieved outstanding predictive metrics: AUC of 96.76%, accuracy (ACC) of 90.49%, MCC of 0.8098, SEN of 90.22%, and SPE of 90.76%.
- Demonstrates robustness and efficiency in identifying RNA ac4C sites.
Conclusions:
- DFM-ac4C is an effective and robust computational tool for RNA ac4C site identification.
- The framework's ability to model multiscale nucleotide interactions enhances prediction accuracy.
- This tool can advance research in RNA modification, gene regulation, and cancer biology.
