Abstract:Abstract Proteins are crucial in regulating every aspect of RNA life, yet understanding their interactions with coding and noncoding RNAs remains limited. Experimental studies are typically restricted to a small number of cell lines and a limited set of RNA-binding proteins (RBPs). Although computational methods based on physico-chemical principles can predict protein-RNA interactions accurately, they often lack the ability to consider cell-type-specific gene expression and the broader context of gene regulatory networks (GRNs). Here, we assess the performance of several GRN inference algorithms in predicting protein-RNA interactions from single-cell transcriptomic data, and propose a pipeline, called scRAPID (single-cell transcriptomic-based RnA Protein Interaction Detection), that integrates these methods with the catRAPID algorithm, which can identify direct physical interactions between RBPs and RNA molecules. Our approach demonstrates that RBP–RNA interactions can be predicted from single-cell transcriptomic data, with performances comparable or superior to those achieved for the well-established task of inferring transcription factor–target interactions. The incorporation of catRAPID significantly enhances the accuracy of identifying interactions, particularly with long noncoding RNAs, and enables the identification of hub RBPs and RNAs. Additionally, we show that interactions between RBPs can be detected based on their inferred RNA targets. The software is freely available at https://github.com/tartaglialabIIT/scRAPID.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is the prediction of protein - RNA interactions. Specifically, the authors focus on how to use single - cell transcriptome data to predict protein - RNA interactions. Currently, experimental methods are usually limited to a small number of cell lines and a limited set of RNA - binding proteins (RBPs), and the knowledge based on CLIP - Seq data is also limited due to limitations in sensitivity, specificity, and reproducibility. Although computational methods based on physicochemical principles can accurately predict protein - RNA interactions, they often lack the ability to consider the gene expression of specific cell types and the context of a broader gene regulatory network (GRNs). To solve these problems, the authors evaluated the performance of several GRN inference algorithms in predicting protein - RNA interactions using single - cell transcriptome data and proposed a pipeline named scRAPID (RNA - Protein Interaction Detection Based on Single - Cell Transcriptome). This pipeline integrates these methods with the cat RAPID algorithm, which can identify the direct physical interactions between RBPs and RNA molecules. Through this method, the authors demonstrated the feasibility of predicting RBP - RNA interactions from single - cell transcriptome data, and its performance can be comparable to or better than the established task of inferring transcription factor - target interactions. In addition, they also proved that by integrating cat RAPID, the accuracy of identifying long non - coding RNA interactions can be significantly improved, and hub RBPs and RNAs can be identified. Finally, the authors also showed that based on the overlap of inferred RNA targets, the interactions between RBPs can be detected. Overall, this study aims to provide a new and scientifically elegant method for predicting protein - RNA interactions from single - cell transcriptome data. By introducing cat RAPID and conducting extensive validation, the inference performance is significantly improved, providing valuable insights into RBP - lncRNA interactions, hub identification, and direct RBP - RBP interactions.

Prediction of protein-RNA interactions from single-cell transcriptomic data

A Hybrid Prediction Method for Plant lncRNA-Protein Interaction.

Discovering putative peptides encoded from non-coding RNAs in ribosome profiling data of Arabidopsis thaliana.

RPI-Pred: predicting ncRNA-protein interaction using sequence and structural information

Predicting Dynamic Cellular Protein–rna Interactions by Deep Learning Using in Vivo RNA Structures

Meta-analysis of RNA interaction profiles of RNA-binding protein using the RBPInper tool

Efficient Framework For Predicting Ncrna-Protein Interactions Based On Sequence Information By Deep Learning

In silico characterization and prediction of global protein–mRNA interactions in yeast

SPOT-Seq-RNA: Predicting Protein-Rna Complex Structure and RNA-binding Function by Fold Recognition and Binding Affinity Prediction.

Recent methodology progress of deep learning for RNA–protein interaction prediction

Computational Analysis of RNA-Protein Interactions via Deep Sequencing.

RBPBind: Quantitative Prediction of Protein-RNA Interactions

RNA-Protein Interaction Prediction Based on Deep Learning: A Comprehensive Survey

Prediction of RNA binding proteins comes of age from low resolution to high resolution.

Discovering RNA-Protein Interactome by Using Chemical Context Profiling of the RNA-Protein Interface

De Novo Prediction of RNA-protein Interactions from Sequence Information.

Structure-Based Prediction of Rna-Binding Domains and Rna-Binding Sites and Application to Structural Genomics Targets

Systematic identification of post-transcriptional regulatory modules

Prediction and Validation of the Unexplored RNA‐binding Protein Atlas of the Human Proteome

Identification of RNA-binding Sites in Proteins by Integrating Various Sequence Information.

Predicting Circrna-Rbp Interaction Sites Using a Codon-Based Encoding and Hybrid Deep Neural Networks