Identify RNA-associated subcellular localizations based on multi-label learning using Chou's 5-steps rule.

Area of Science:

• Molecular Biology
• Bioinformatics
• Computational Biology

Background:

• Cellular functions depend on biomolecule localization within specific compartments.
• RNA molecules are precisely located to enable diverse cellular processes.
• Existing RNA localization classifiers primarily focus on single-label classification, limiting their scope.

Purpose of the Study:

• To address the limitation of single-label classification in RNA subcellular localization.
• To develop a robust method for multi-label RNA subcellular localization prediction.
• To establish comprehensive human RNA subcellular localization datasets.

Main Methods:

• Extraction of multi-label RNA subcellular localization datasets across various RNA types.
• Development of nucleotide property composition models for feature extraction.
• Application of multiple kernel learning based on Hilbert-Schmidt independence criterion for information fusion.
• Integration with a support vector machine model for multi-label classification.

Main Results:

• Construction of multi-label RNA subcellular localization datasets for four RNA categories and human RNAs.
• Effective feature extraction using nucleotide property composition models.
• Successful fusion of multivariate information via multiple kernel learning.
• Achieved average precision scores of 0.703, 0.757, 0.787, and 0.800 on four RNA datasets.

Conclusions:

• The novel method demonstrates superior performance compared to existing prediction tools on benchmark datasets.
• The developed method accurately identifies multi-label RNA subcellular localizations.
• A user-friendly web server has been established for practical application of the method.