Practical considerations on performing and analyzing CLIP-seq experiments to identify transcriptomic-wide RNA-protein interactions

Abstract

RNA-binding proteins are important players in post-transcriptional regulation, such as modulating mRNA splicing, translation, and degradation under diverse biological settings. Identifying and characterizing the RNA substrates is a critical step in deciphering the function and molecular mechanisms of the target RNA-binding proteins. High-throughput sequencing of the RNA fragments isolated by crosslinking immunoprecipitation (CLIP-seq) is one of the standard techniques to identify the in vivo transcriptome-wide binding sites of the target RNA-binding protein. This method is widely used in functional and mechanistic characterizations of RNAbinding proteins. In this review, we provide several practical considerations on performing and analyzing CLIP-seq experiments. Particularly, we focus on how to perform CLIP-seq experiments on endogenous RNA-binding proteins. In addition, we provide a practical summary on how to choose and use computational pipelines from an increasing number of computational methods and packages that are available for analyzing the sequencing datasets from the CLIP-seq experiments. We hope these practical considerations will facilitate experimental biologists in performing and analyzing CLIP-seq experiment to obtain biologically relevant mechanistic insights.

nearBynding: A flexible pipeline characterizing protein binding to local RNA structure

“…The dampened prediction accuracy of these RBPs is perhaps due to biological confounders, such as subcellular localization or cofactors that affect binding preference, which cannot be discerned from CLIP data alone. The predictive power of these state-ofthe-art algorithms may be limited by their reliance exclusively on sequence-based RNA structure prediction and their lack of accommodation for experimentally-derived RNA structure information (reviewed [22,23]). A further limitation of these algorithms is that they predict RNA structures only from short intervals [22].…”

“…The predictive power of these state-ofthe-art algorithms may be limited by their reliance exclusively on sequence-based RNA structure prediction and their lack of accommodation for experimentally-derived RNA structure information (reviewed [22,23]). A further limitation of these algorithms is that they predict RNA structures only from short intervals [22]. Although more computationally efficient, this approach sacrifices the possibility of long-range contacts that affect secondary structure or of tertiary conformations, which may shield or expose binding sites.…”

Breeding for Climate Change Resilience: A Case Study of Loblolly Pine (Pinus taeda L.) in North America

“…This has been a notorious bottleneck in the study of AS and APA in non-model plants, particularly woody species with complex genomes. Chen et al (2020) compiled recent publications focused on AS studies in different tree species at various stages of development and in response to various stresses. The group, highlighted major contributions in the study of AS in woody species that have been focused on development and stress-responses, including fruit ripening (Mica et al, 2010;Gupta et al, 2015), flower morphogenesis (Ai et al, 2012), wood formation (Xu et al, 2014), drought stress (Guo et al, 2017;Ding et al, 2020), and cold stress (Xiao and Nassuth, 2006;Rahman et al, 2014;Li et al, 2020).…”

“…The group, highlighted major contributions in the study of AS in woody species that have been focused on development and stress-responses, including fruit ripening (Mica et al, 2010;Gupta et al, 2015), flower morphogenesis (Ai et al, 2012), wood formation (Xu et al, 2014), drought stress (Guo et al, 2017;Ding et al, 2020), and cold stress (Xiao and Nassuth, 2006;Rahman et al, 2014;Li et al, 2020). Strategies to advance in the functional analysis of AS in woody species have been proposed (Chen et al, 2020). These are some examples of the applicability of these methods in angiosperms: phylogenetic analysis and spatial expression analysis to unravel functional conserved genes and cis-elements involved in AS (Li et al, 2017;Liu et al, 2017); identification of regulatory genes in signaling cascades that exhibit AS in specific stress-depend patterns (Li Z. et al, 2019) specifically induced by temperature stress (cold and heat) (Palusa et al, 2007).…”

Spatial correlation statistics enable transcriptome-wide characterization of RNA structure binding

“…Algorithms such as GraphProt and iDeepS incorporate a post-processing step to easily visualize RBP sequence and structure preferences ( Maticzka et al., 2014 ; Pan and Shen, 2018 ), but these algorithms only provide visualization of structure information for a short binding motif (7–12 nucleotides). The predictive power of many state-of-the-art algorithms may be limited by their reliance exclusively on sequence-based RNA structure prediction and their lack of accommodation for experimentally derived RNA structure information (reviewed by Chen et al., 2019 ; Sasse et al., 2018 ). The recent algorithm PrismNet has begun addressing these problems by allowing the incorporation of in vivo click selective 2′-hydroxyl acetylation and profiling experiment (icSHAPE) data ( Sun et al., 2021 ), but all algorithms to resolve structure-based RBP motifs still rely exclusively on analysis of overlapping coordinates and do not offer insight about the RNA structure surrounding those motifs, despite evidence that such context can be important in RBP binding ( Carlile et al., 2019 ; Jarmoskaite et al., 2019 ).…”