Characterization and identification of long non-coding RNAs based on feature relationship

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Long non-coding RNAs (lncRNAs) play crucial roles in biological processes and diseases.
• Computational identification of lncRNAs is challenging due to limited high-quality data for many species.
• Existing methods often require species-specific training data or prior sequence knowledge.

Purpose of the Study:

• To characterize the distinct feature relationships between lncRNAs and protein-coding RNAs.
• To develop a novel, cross-species algorithm for accurate lncRNA identification.
• To provide a computational tool that does not require prior knowledge or species-specific data.

Main Methods:

• Characterization of lncRNAs versus protein-coding RNAs based on feature relationships, specifically open reading frame length and GC content.
• Development of the LGC algorithm leveraging these distinct feature relationships.
• Validation of LGC performance on large-scale empirical datasets across diverse species.

Main Results:

• A universal divergence in the relationship between open reading frame length and GC content was identified between lncRNAs and protein-coding RNAs.
• The LGC algorithm accurately distinguishes lncRNAs from protein-coding RNAs in a cross-species manner.
• LGC demonstrated superior performance compared to existing algorithms, achieving >90% accuracy with well-balanced sensitivity and specificity.

Conclusions:

• The study presents a novel method for characterizing and computationally identifying lncRNAs based on fundamental feature relationships.
• LGC offers a robust and accurate cross-species solution for lncRNA identification, overcoming previous limitations.
• LGC has broad potential utility for computational analysis of lncRNAs across a wide range of species.