Skip to content

Signatures of mutational processes in human cancer

Why this mattered

Before this paper, cancer genomes were usually read as lists of driver mutations plus background noise. Alexandrov and colleagues made the “background” interpretable: by analyzing millions of somatic mutations across thousands of tumors, they showed that cancers carry reproducible mutational signatures, each reflecting the history of DNA damage, repair failure, replication error, or enzymatic editing that shaped the tumor. The paradigm shift was to treat the cancer genome not only as a catalogue of selected lesions, but also as a molecular archive of past mutational processes.

What became newly possible was systematic forensic inference from tumor genomes. The paper extracted more than 20 signatures across 30 cancer types, connecting some to known causes such as ultraviolet light, tobacco exposure, age-related processes, APOBEC cytidine deaminase activity, and DNA maintenance defects, while leaving others as cryptic signals for future work. It also placed localized hypermutation, kataegis, into a pan-cancer framework. This converted whole-genome sequencing from a mutation-finding exercise into a quantitative way to ask which biological processes had acted in each cancer and how strongly.

Subsequent cancer genomics built directly on this idea. Mutational signature analysis became a standard layer of tumor interpretation, later expanded through COSMIC signature catalogs, PCAWG-scale analyses, clock-like mutation studies, homologous-recombination-deficiency signatures, mismatch-repair diagnostics, and therapy-resistance investigations. Its influence is visible in both basic and translational oncology: signatures now help reconstruct cancer etiology, stratify tumors by repair defects, identify exposure histories, and guide hypotheses about prevention or treatment sensitivity.

Abstract

(no abstract available)

Sources