IN CANCER CELLS THE GENETIC CODE HAS, by definition, become corrupted somewhere along the line. But a major challenge in cancer genomics has been identifying which mutations are the actual “drivers” of cancer, responsible for unimpeded growth of the cell, and which are just “passengers.” A typical cancer genome will have only five to 10 driver mutations but thousands or even millions of passenger mutations along for the ride.

Researchers tend to classify recurrent mutations—those that have been found time and again in the tumors of many different patients—as driver mutations. But new research published in Science reveals that some mutations that occur with great frequency are, in fact, dense groups of passengers—passenger hotspots—that come about because of a quirk of cell replication.

One human enzyme, APOBEC3A, has a strong preference for DNA structures called stem-loops, or hairpins, which occur during DNA replication. The stretches of DNA that form hairpins may mutate up to 200 times as frequently as non-hairpin sites, and the mutations occur in a dense group.

The authors note that these passenger hotspot mutations owe their prevalence in cancer cells to their attractiveness to APOBEC3A and “not to any effects on tumor fitness.” These findings highlight the importance of being able to recognize passengers—and not to waste therapies or drug development resources on these false culprits in cancer cells.