Shiyong Liu

PRIME-BSPre predicts genome-wide protein-RNA binding sites using templates that integrate RNA sequence, secondary, and tertiary structures, showing strong performance on human RBP datasets.

CPPred-sORF improves coding potential prediction for sORFs by using non-AUG start codons and achieves high sensitivity and specificity.

The paper introduces RMalign, a size-independent RNA 3D structural alignment tool, which improves benchmarking performance over existing methods and updates the PRIME RNA-protein modeling approach.

Deep-RBPPred predicts RNA binding proteins across multiple species using deep learning, achieving high accuracy and speed compared to existing methods.