JUM is a computational method for comprehensive annotation-free analysis of alternative pre-mRNA splicing patterns

Significance Alternative pre-mRNA splicing (AS) is a critical gene regulatory mechanism to produce diverse, tissue-specific, and functionally distinct protein profiles in eukaryotes to maintain normal cellular functions. Aberrant AS patterns are constantly associated with many human diseases, including cancer. The exceptional complexity of AS imposes a major challenge to analyzing AS across various tissues and cell types. Here we present a computational algorithm to profile and quantitate tissue-specific AS profiles from RNA-sequencing data without any prior knowledge of the host transcriptome. The junction usage model shows consistent superior performance in both specificity and sensitivity compared with other currently available AS analysis methods, and can be readily applied to a wide range of RNA samples from different organisms for accurate and comprehensive analyses of AS.