Scientists have long known that retroviruses pose a risk because they can lodge in or near a cancer-causing gene.

As a result of implanting a particular cancer-causing gene, half the female offspring of the genetically altered mice will develop cancer.

But the capture of such cancer-causing genes was considered a freak event, a mistake not necessarily beneficial to the virus.

These, in turn, enable DNA to resist the action of cancer-causing genes.

But scientists also knew there was a theoretical risk that a retrovirus would lodge near a cancer-causing gene and turn it on.

The discovery of bcl-2 in 1985 was a result of the vigorous search for cancer-causing genes in the 1980's.

Many people are insured through their employers as part of large groups, and they would be covered even if they did have a cancer-causing gene.

If areas around oncogenes become unmethylated these cancer-causing genes have the potential to be transcribed at an alarming rate.

Algorithm CanPredict was used to indicate how closely a specific gene resembles known cancer-causing genes.

The cultured mouse cells that are the universal test for cancer-causing genes work only because they are very far from normal.