Résumé : Cancer is one of the leading causes of human death. Many efforts have made to understand its mechanism and have further identified many proteins and DNA sequence variations as suspected targets for therapy. However, drugs targeting these targets have low success rates, suggesting the basic mechanism still remains unclear. Here, we develop a computational software combining Cox proportional-hazards model and stability-selection to unearth an overlooked, yet the most important cancer drivers hidden in massive data from The Cancer Genome Atlas (TCGA), including 11,574 RNAseq samples and clinic data. Generally, noncoding RNAs primarily regulate cancer deaths and work as the deadliest cancer inducers and repressors, in contrast to proteins as conventionally thought. Especially, processed-pseudogenes serve as the primary cancer inducers, while lincRNA and antisense RNAs dominate the repressors. Strikingly, noncoding RNAs serves as the universal strongest regulators for all cancer types although personal clinic variables such as alcohol and smoking significantly alter cancer genome. Furthermore, noncoding RNAs also work as central hubs in cancer regulatory network and as biomarkers to discriminate cancer types. Therefore, noncoding RNAs overall serve as the deadliest cancer regulators, which refreshes the basic concept of cancer mechanism and builds a novel basis for cancer research and therapy. Biological functions of pseudogenes have rarely been recognized. Here we reveal them as the most important cancer drivers for all cancer types from big data, breaking a wall to explore their biological potentials.