NAViGaTing the micronome--using multiple microRNA prediction databases to identify signalling pathway-associated...

Area of Science:

• Molecular Biology
• Bioinformatics

Background:

• MicroRNAs (miRNAs) are small RNAs regulating gene expression at transcript or protein levels.
• Existing databases predict miRNA:target relationships using diverse algorithms.
• Accurate miRNA:target interaction data is crucial for understanding biological processes.

Purpose of the Study:

• To integrate miRNA prediction databases for accurate miRNA:target relationship elucidation.
• To model the miRNA:transcript interactome (micronome) for pathway and disease association studies.
• To develop a user-friendly portal (mirDIP) for accessing integrated miRNA data.

Main Methods:

• Integrated multiple miRNA prediction databases.
• Compared predictions with in vitro data.
• Utilized cross-database predictions to build the micronome model.
• Employed NAViGaTOR for constructing interaction networks based on literature, KEGG, and Reactome pathways.

Main Results:

• The mirDIP portal integrates prediction databases to identify accurate miRNA:target relationships.
• Pathway-based networks show significantly higher miRNA involvement than expected by chance, indicating co-targeting.
• Identified two distinct miRNA classes: universe miRNAs (involved in many pathways) and intra-pathway miRNAs (targeting genes within a single pathway).
• Universe miRNAs exhibit more targets, are more studied, and have higher connectivity in cancer pathways compared to intra-pathway miRNAs.

Conclusions:

• Pathway analysis of mirDIP data confirms miRNA involvement in intra-pathway signaling.
• The existence of universe and intra-pathway miRNAs suggests a hierarchical organization of miRNA regulation.
• Differential involvement of these miRNA classes at the disease level is implied.