RPI-SE: a stacking ensemble learning framework for ncRNA-protein interactions prediction using sequence information.

Area of Science:

• Bioinformatics
• Computational Biology
• Molecular Biology

Background:

• Non-coding RNAs (ncRNAs) and proteins are crucial in biological processes.
• Experimental methods for detecting ncRNA-protein interactions are often slow and costly.
• Computational approaches offer efficient alternatives for studying these interactions.

Purpose of the Study:

• To develop an accurate and efficient computational framework for predicting ncRNA-protein interactions.
• To leverage sequence information for improved prediction accuracy.
• To provide a valuable tool for advancing biomedical research.

Main Methods:

• Developed RPI-SE, a stacking ensemble computational framework.
• Utilized Position Weight Matrix and Legendre Moments for protein sequence feature extraction.
• Employed k-mer sparse matrix for ncRNA sequence feature extraction.
• Integrated diverse base classifiers within an ensemble learning framework.

Main Results:

• RPI-SE demonstrated high accuracy and robustness in predicting ncRNA-protein interactions.
• Performance was validated on three benchmark datasets using five-fold cross-validation.
• The framework outperformed existing state-of-the-art methods.

Conclusions:

• RPI-SE is a competent tool for ncRNA-protein interaction prediction.
• The framework offers high accuracy and robustness.
• This computational tool can significantly advance ncRNA-protein interaction research.