IDDLncLoc: Subcellular Localization of LncRNAs Based on a Framework for Imbalanced Data Distributions

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Long non-coding RNAs (lncRNAs) are vital regulators in cellular processes, including gene expression and protein regulation.
• Experimental determination of lncRNA subcellular localization is challenging, costly, and difficult to reproduce.
• Accurate lncRNA localization is crucial for understanding their diverse biological functions.

Purpose of the Study:

• To develop a novel computational method for predicting the subcellular localization of long non-coding RNAs (lncRNAs).
• To address the limitations of experimental methods in lncRNA localization studies.
• To provide an accurate and efficient tool for lncRNA subcellular localization prediction.

Main Methods:

• An ensemble model, IDDLncLoc, was developed for predicting lncRNA subcellular localization.
• Sequence encoding involved dinucleotide-based auto-cross covariance, k-mer composition, and composition, transition, and distribution (CTD) features.
• Feature selection utilized binomial distribution and recursive feature elimination.
• Data imbalance was managed using mini-batch oversampling, random sampling, and stacking ensemble strategies.

Main Results:

• IDDLncLoc achieved a high accuracy of 94.96% on the benchmark dataset for lncRNA subcellular localization.
• The proposed method outperformed existing state-of-the-art methods by 2.59%.
• The results confirm the efficacy and superiority of IDDLncLoc in predicting lncRNA subcellular localization.

Conclusions:

• IDDLncLoc provides a robust and accurate computational solution for predicting lncRNA subcellular localization.
• The method overcomes the complexities and costs associated with experimental approaches.
• A user-friendly web server is available, facilitating broader research applications in lncRNA biology.