WAFNRLTG: A Novel Model for Predicting LncRNA Target Genes Based on Weighted Average Fusion Network Representation...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Long non-coding RNAs (lncRNAs) are crucial regulators of gene expression, despite not encoding proteins.
• Dysregulation of lncRNAs is implicated in various complex human diseases.
• Accurate prediction of lncRNA target genes is essential for understanding lncRNA functions.

Purpose of the Study:

• To develop a novel computational model for predicting lncRNA target genes.
• To overcome limitations of traditional prediction methods.
• To improve functional annotation of lncRNAs.

Main Methods:

• Constructed a heterogeneous network integrating lncRNA/mRNA sequence similarity, lncRNA-mRNA, lncRNA-miRNA, and mRNA-miRNA interaction networks.
• Employed network representation learning to obtain node vectors.
• Utilized weighted average fusion and an XGBoost classifier to predict lncRNA target genes.

Main Results:

• The proposed WAFNRLTG model achieved high performance with AUC scores of 0.9410 and 0.9350.
• AUPR scores reached 0.9391 and 0.9350 on training and independent datasets, respectively.
• Case studies validated the model's effectiveness in predicting target genes for common lncRNAs.

Conclusions:

• WAFNRLTG demonstrates superior accuracy in predicting lncRNA target genes.
• The model provides a valuable tool for advancing lncRNA research and disease mechanism studies.
• The developed method and data are publicly available for broader scientific use.