Unraveling sequence determinants which drive protein-RNA interaction is crucial for studying binding mechanisms and the impact of genomic variants. RBPNet, a novel deep learning method, predicts CLIP crosslink count distribution from RNA sequence at single-nucleotide resolution. RBPNet achieves high generalization on eCLIP, iCLIP and miCLIP assays, outperforming state-of-the-art classifiers. Via model-intrinsic bias correction, RBPNet identifies predictive RNA sub-sequences corresponding to known binding motifs and enables variant-impact scoring via in silico mutagenesis. Together, RBPNet improves inference of protein-RNA interaction, as well as mechanistic interpretation of predictions.
Author: Marc Horlacher.