Abstract 3139: Pan-cancer splicing analysis reveals shared drivers of malignant transformation and survival

Abstract Alternative splicing is one of the primary mechanisms used to achieve mRNA transcript and proteomic diversity in higher order eukaryotes. In cancer, altered mRNA splicing can lead to aberrant protein products that promote oncogenic transformation, metastasis and confer chemotherapy resistance, due to splicing factor mutation or mis-expression. We hypothesize that the role of splicing imbalance in cancer is grossly underestimated and is likely regulated by the same recurrent global disruptions observed across human cancers. To determine common and malignancy specific splicing subtypes across cancer we designed a novel integrated computational workflow that uses genome variant data from RNA-Seq in conjunction with fully unsupervised analyses (NMF and SVM) to identify novel patient splicing-defined subtypes (OncoSplice). Applied to 20 adult and pediatric cancers with large cohorts, we identified common-recurrent splicing subtypes associated with MYC-hyperactivation and TGF-beta signaling, lineage reprogramming, tumor infiltration and broad mutation impacts. Such recurrent and tumor specific subtypes were frequently associated with poor prognosis and novel dominantly regulated driving splicing events (e.g., transcription factors). Broad splicing subtypes were associated with circadian dysregulation, as determined from normal healthy tissues (GTEx) and in many cases were found to phenocopy well-described splicing mutation impacts or were associated with new RNA-binding proteins (RBPs) evidenced by orthogonal computational predictions. Splicing subtype associated RBPs were frequently undergo autoregulation, as further evidenced by CLIP-Seq and RBP knockdown. To enhance the RNA community’s ability to explore hundreds of known and novel splicing variation across cancers and healthy tissues, we provide an interactive online splicing explorer at oncosplice.org. Together, these data highlight previously unknown regulatory relationships and prognostic associations in cancer associated with broad and targeted splicing regulation. Citation Format: Anukana Bhattacharjee, Audrey Crowther, Guangyuan Li, Meenakshi Venkatasubramanian, Dan Schnell, Stuart Hay, Preeti Singh, Krithika R. Subramanian, Kashish Chetal, Xiaoting Chen, Aishwarya Kulkarni, Matthew T. Weirauch, Nathan Salomonis. Pan-cancer splicing analysis reveals shared drivers of malignant transformation and survival [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3139.