miRmap

miRmap predicts miRNA-mediated mRNA repression strength by integrating thermodynamic, evolutionary, probabilistic, and sequence-based features to rank and prioritize mRNA targets.

Key Features:

• Comprehensive prediction approaches: Integrates thermodynamic, evolutionary, probabilistic, and sequence-based features to evaluate mRNA repression by miRNAs.
• Eleven predictor features (three novel): Incorporates 11 predictor features, including three novel contributions, to enhance predictive accuracy of miRNA–mRNA interactions.
• Feature correlation and comparison: Enables unbiased comparison of feature correlations using high-throughput experimental data from immunopurification, transcriptomics, proteomics, and polysome fractionation.
• Target site accessibility: Identifies target site accessibility as the most predictive indicator of miRNA-mediated mRNA repression strength.
• PhyloP-based evolutionary feature: Uses a PhyloP-based feature to evaluate negative selection, which is the best-performing predictor within the evolutionary category.
• Integrated predictive model: Combines all predictive features into an integrated model that improves prediction accuracy relative to existing models such as TargetScan.

Scientific Applications:

• Prioritization for experimental validation: Ranks and prioritizes candidate miRNA targets for follow-up experimental validation.
• Analysis of miRNA-mediated repression with high-throughput data: Facilitates interpretation and comparison of immunopurification, transcriptomics, proteomics, and polysome fractionation datasets in the context of miRNA repression.
• Benchmarking predictive features: Identifies and compares the most predictive features of miRNA target repression against existing prediction models such as TargetScan.

Methodology:

Integrates thermodynamic, evolutionary (including a PhyloP-based negative-selection feature), probabilistic, and sequence-based predictor features (11 features, three novel) into an integrated model and compares feature correlations and validation using high-throughput datasets from immunopurification, transcriptomics, proteomics, and polysome fractionation.