Cargando…

PrismNet: predicting protein–RNA interaction using in vivo RNA structural information

Fundamental to post-transcriptional regulation, the in vivo binding of RNA binding proteins (RBPs) on their RNA targets heavily depends on RNA structures. To date, most methods for RBP–RNA interaction prediction are based on RNA structures predicted from sequences, which do not consider the various...

Descripción completa

Detalles Bibliográficos
Autores principales: Xu, Yiran, Zhu, Jianghui, Huang, Wenze, Xu, Kui, Yang, Rui, Zhang, Qiangfeng Cliff, Sun, Lei
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10320048/
https://www.ncbi.nlm.nih.gov/pubmed/37140045
http://dx.doi.org/10.1093/nar/gkad353
Descripción
Sumario:Fundamental to post-transcriptional regulation, the in vivo binding of RNA binding proteins (RBPs) on their RNA targets heavily depends on RNA structures. To date, most methods for RBP–RNA interaction prediction are based on RNA structures predicted from sequences, which do not consider the various intracellular environments and thus cannot predict cell type-specific RBP–RNA interactions. Here, we present a web server PrismNet that uses a deep learning tool to integrate in vivo RNA secondary structures measured by icSHAPE experiments with RBP binding site information from UV cross-linking and immunoprecipitation in the same cell lines to predict cell type-specific RBP–RNA interactions. Taking an RBP and an RNA region with sequential and structural information as input (‘Sequence & Structure’ mode), PrismNet outputs the binding probability of the RBP and this RNA region, together with a saliency map and a sequence–structure integrative motif. The web server is freely available at http://prismnetweb.zhanglab.net.