Veronica F. Busa

RNA-binding proteins (RBPs) play diverse, important roles in cellular processes and regulation. One RBP can bind and affect hundreds of RNAs and cause a cascade effect across the transcriptome. Understanding RBP binding specificity and function is therefore central in understanding human biology and diseases. RBP binding relies on the sequence and folded structure of the RNA. There are many tools to study sequence-based RBP binding preferences, but few for identifying structure-based RBP binding preferences. nearBynding is a pipeline that incorporates RBP binding sites and RNA structure information to discern local RNA structure for regions bound by an RBP. The motivation for this pipeline is three-fold: to visualize RNA structure at and proximal to RBP binding transcriptome-wide; to provide a flexible scaffold to study RBP binding preferences relative to diverse RNA structure data types; and to allow for RBP binding data analysis without extensive pre-processing. nearBynding extends the algorithm StereoGene to allow for direct estimation of correlation among pairs of continuous or interval features along the transcriptome, such as RNA structure and RBP binding. Using both sequence-based RNA structure predictions and experimental RNA annotations, nearBynding can recapitulate published RBP structural binding preferences and observe new binding profiles. nearBynding is efficient enough to run on a personal computer and is available as an easy-to-use R package on the biology software database Bioconductor.