It is increasingly appreciated that alternative splicing plays a key role in generating functional specificity and diversity in cancer. However, the mechanisms by which cancer mutations perturb splicing remain unknown. Here, we developed a network-based strategy, DrAS-Net, to investigate more than 2.5 million variants across cancer types and link somatic mutations with cancer-specific splicing events. We identi-fied more than 40,000 driver variant candidates and their 80,000 putative splicing targets deregulated in 33 cancer types and inferred their functional impact. 

Strikingly, tumors with splicing perturbations show reduced expression of immune system-related genes and increased expression of cell proliferation markers. Tumors harboring different mutations in the same gene often exhibit distinct splicing perturbations. Further stratification of 10,000 patients based on their mutation-splicing relationships identifies subtypes with distinct clinical features, including survival rates. Our work reveals how single-nucleotide changes can alter the repertoires of splicing isoforms, providing insights into oncogenic mechanisms for precision medicine.

The paper by Li et al can be found here.