RBPsuite

RBPsuite predicts RNA-binding protein (RBP) binding sites on linear RNAs and circular RNAs (circRNAs) using deep learning to identify potential RBP–RNA interactions.

Key Features:

• Deep Learning-Based Predictions: Employs deep learning models to predict RBP binding sites and compute binding scores.
• Support for Linear and Circular RNAs: An updated version of iDeepS is used for linear RNAs and CRIP is used for circular RNAs (circRNAs).
• Segmented Analysis: Divides input RNA sequences into 101-nucleotide segments to assess interactions between segments and RBPs.
• Motif Detection and Scoring Distribution: Detects verified motifs within binding segments and provides a distribution of binding scores along the full-length RNA sequence.

Scientific Applications:

• Gene regulation studies: Supports investigation of RBP roles in gene regulation by identifying putative binding sites.
• RNA stability and localization: Enables analysis of RBP interactions that may affect RNA stability and subcellular localization.
• Splicing and translation regulation: Facilitates study of RBP-mediated effects on splicing and translation.
• circRNA research: Facilitates identification of RBP binding sites on circular RNAs where traditional prediction methods may be less effective.

Methodology:

Input RNA sequences are divided into 101-nucleotide segments; deep learning models are applied to predict segment–RBP interactions using an updated iDeepS for linear RNAs and CRIP for circRNAs, with motif detection and generation of binding score distributions along full-length RNAs.