iPseU-NCP: Identifying RNA pseudouridine sites using random forest and NCP-encoded features

Area of Science:

• Biochemistry and Molecular Biology
• Bioinformatics and Computational Biology

Background:

• Pseudouridine modification is a prevalent RNA modification in both prokaryotes and eukaryotes, occurring in various RNA types like mRNA, rRNA, and tRNA.
• Understanding pseudouridine modification is crucial for advancing drug discovery and gene therapies.
• Current laboratory methods for pseudouridine identification are often costly and require specialized expertise.

Purpose of the Study:

• To develop an efficient computational framework for predicting pseudouridine sites in RNA sequences.
• To introduce iPseU-NCP, a novel method utilizing the Random Forest algorithm and nucleotide chemical properties.

Main Methods:

• The iPseU-NCP framework was developed using the Random Forest (RF) algorithm.
• Nucleotide chemical properties (NCP) were extracted from RNA sequences to serve as features.
• Performance was evaluated on a benchmark dataset from Chen et al. (2016) and compared against existing state-of-the-art methods.

Main Results:

• iPseU-NCP demonstrated significant improvements in Matthew's Correlation Coefficient (MCC) compared to iPseU-CNN, PseUI, and iRNA-PseU on both H. sapiens and S. cerevisiae datasets.
• The model achieved higher accuracy rates across independent test datasets when compared to the other three methods.
• iPseU-NCP exhibited superior performance across most evaluation metrics, highlighting its stability and effectiveness.

Conclusions:

• The iPseU-NCP framework, integrating RF and NCP-encoded features, surpasses existing methods in identifying pseudouridine sites.
• This computational approach offers a promising strategy for addressing biological challenges associated with pseudouridine modifications and human diseases.