pirnaPre

pirnaPre predicts potential targets of PIWI-interacting RNAs (piRNAs) on messenger RNAs (mRNAs) by applying a support vector machine (SVM) model using features derived from Miwi CLIP-Seq and position-derived information to identify protein-coding genes under piRNA-mediated regulation.

Key Features:

• Machine Learning Approach: Employs a support vector machine (SVM) classifier integrating Miwi CLIP-Seq-derived features and position-derived features to predict piRNA targets on mRNAs.
• Data Integration: Combines experimental Miwi CLIP-Seq interaction data with positional sequence/context features to inform target prediction.
• Systematic Prediction: Produces comprehensive lists of candidate mRNA targets for piRNAs based on the combined feature set.
• Validation and Support: Predictions are supported by reanalysis of microarray datasets showing significant upregulation of 2587 protein-coding genes predicted as piRNA targets following disruption of Miwi's slicer activity.

Scientific Applications:

• Gene Expression Regulation: Facilitates investigation of piRNA contributions to gene expression regulation beyond transposable element (TE) silencing, including roles in the germline.
• Genome Integrity Maintenance: Helps elucidate mechanisms by which piRNAs contribute to genome integrity through mRNA targeting.
• Non-coding RNA Research: Supports studies in non-coding RNA biology by prioritizing candidate piRNA–mRNA interactions for experimental validation.

Methodology:

Features were extracted from Miwi CLIP-Seq data and position-derived sequence/context features, and a support vector machine (SVM) classifier was trained on these features with validation against empirical microarray data.