Answers are grounded in this paper and cite their sources.
Cancer is driven by alterations across multiple molecular layers, but whether multi-omics drivers carry prognostic weight has received little systematic attention. DriverDBv5 (https://driverdb.bioinfomics.org/) extends DriverDB from driver identification to prognostic evaluation. It has been expanded to 187 cohorts, 43 862 patients, and 129 655 samples, with 154 cohorts from non-TCGA (The Cancer Genome Atlas) repositories reconciled by a generalized multi-omics barcode. Driver genes can be evaluated with Cox and mixture cure models, with machine learning on the full multi-omics feature matrix, and for the joint prognostic effect of alterations in two omics layers. The Customized Analysis module has grown from 4 to 10 tools for user-defined gene sets, subgroups, and covariates, and a human body map offers an anatomical entry point. DriverDBv5 thus reveals not only which alterations drive a tumor, but also which of them relate to patient outcome.
No takes yet. Share an insight, caveat, or question.
Lai et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: