ceRNAR

ceRNAR identifies competitive endogenous RNA (ceRNA)–miRNA triplets by accounting for miRNA expression levels to detect ceRNA regulatory events.

Key Features:

• Identification of ceRNA pairs: Employs a novel rank-based correlation approach that accounts for miRNA influence using a running-sum correlation statistic to detect candidate ceRNA pairs.
• Sample clustering and peak merging: Performs sample clustering based on gene–gene correlations using circular binary segmentation followed by peak merging to pinpoint recurrent patterns across samples.
• Downstream analyses: Provides network analysis, functional annotation, survival analysis, and external validation for biological interpretation of identified triplets.
• Integration capabilities: Supports integration with other bioinformatics tools to combine results with external datasets and analyses.
• Method validation and real-data results: Validated via simulation studies showing high sensitivity, low false-positive rates, and efficient computational time, and applied to lung cancer datasets where it identified the MAP4K3 ceRNA bridged by hsa-let-7c-5p.

Scientific Applications:

• Gene regulation studies: Enables identification and analysis of ceRNA–miRNA interactions to investigate post-transcriptional regulatory networks.
• Cancer research and biomarker discovery: Facilitates discovery of disease-associated ceRNA–miRNA triplets and candidate targets, exemplified by findings in lung cancer such as the MAP4K3–hsa-let-7c-5p triplet.

Methodology:

Uses a rank-based correlation approach with a running-sum correlation statistic, circular binary segmentation for sample clustering, peak merging, and downstream modules for network analysis, functional annotation, survival analysis, and external validation; validated by simulation studies and applied to lung cancer datasets.