Abstract
As a molecular feature that characterizes most tumor types, cancer-associated splicing dysregulation largely arises from recurrent genetic mutations and altered expression of trans-acting splicing factors. Although splicing factor mutations occur less frequently in solid tumors, splicing disorders are pervasive and proven to promote tumorigenesis. However, it still lacks an efficient way to identify the key splicing factors at the top regulatory hierarchy whose abnormal expressions induce such splicing disorders and drive phenotypic variability. Here, MRAS (Master Regulator analysis of Alternative Splicing) is introduced, a computational method designed to pinpoint the pivotal splicing factors that play a central role in shaping splicing regulatory networks and influencing cellular processes. MRAS is demonstrated its power by identifying master splicing regulators associated with various disease phenotypes, including tumor initiation, progression, metastasis, and treatment resistance. Moreover, by applying MRAS to single-cell RNA-seq data, crucial regulatory relationships that govern cell-type specific splicing programs have been uncovered. Overall, MRAS presented as an accurate and versatile approach to unraveling the underlying mechanism of splicing regulation.