The miniature machine, which uses a handful of DNA molecules as its software, measures just 100 nanometres across, meaning that a thousand billion would fit into a teardrop and almost ten million on the full stop at the end of this sentence.

Despite its tiny dimensions, it is sufficiently powerful to detect the chemical signature of prostate or lung cancer cells, and then release a drug to destroy them. The entire process takes just a couple of minutes.

The breakthrough, at the Weizmann Institute in Israel, offers the strongest indication yet that it will eventually be possible to build tiny medical “nanosubs” that hunt down tumours and germs before delivering their drugs.

Although such a “smart drug” or “doctor in a cell” is decades away, the prospect is considered among the most exciting of all the medical applications of nanotechnology.

It would allow doctors to target disease much more precisely and completely than before. Nanosubs could potentially check every cell in the body for cancer, knocking out the abnormal ones while leaving healthy ones alone. Tumours would be eliminated more thoroughly, while patients would be spared the unpleasant side-effects of chemotherapy, which poisons normal as well as cancerous tissue.

The idea of using a microscopic hunter-killer submarine to treat disease has echoes of the 1960s science fiction film Fantastic Voyage, in which a vessel crewed by Raquel Welch and Stephen Boyd is shrunk by military researchers and dispatched to destroy a blood clot threatening the life of a key scientist.

In the Weizmann study, details of which were published online yesterday by the journal Nature, Ehud Shapiro led a team which adapted a miniature biological computer first built last year and recognised by The Guinness Book of World Records as the smallest in the world.

“Our medical computer might one day be administered as a drug, and be distributed throughout the body by the bloodstream to detect disease markers autonomously and independently in every cell,” Professor Shapiro said.

“In this way, a single cancer cell could be detected and destroyed before the tumour develops. Even in a late-stage cancer, this kind of treatment could reach every secondary growth, however small, and effectively terminate the disease.

The system’s input is a single double-stranded DNA molecule, which assesses the concentration of RNA (an indicator of genetic activity) molecules around it. These can be overproduced or underproduced by different types of cancer cell. This information is then fed into the computer’s computation module, which combines DNA software with hardware composed of an enzyme known as FokI. It examines the RNA signals picked up by the input molecule to determine the diagnosis.

If cancer is detected, the computer orders the release of a single-strand DNA molecule designed to induce cancer cells to self-destruct. The whole assembly comprises no more than 20 molecules, which are held together in solution rather than connected to one another like the components of a laptop.

In the experiments, the computer successfully identified both prostate and lung cancer, and released appropriate drugs to fight them. These results, however, were obtained in a highly controlled test tube environment, in which the computer was not exposed to bodily proteins, acids and fats that might destroy it or disrupt its function.

Professor Shapiro said that the technology would have to be refined considerably before it could be tested in animal models, let alone human beings. Nevertheless, the research “represents the first proof of concept and the first demonstration of a possible real-life application for this kind of computer”. He said: “It is clear that the road to realising our vision is a long one. It may take decades before such a system operating inside the human body becomes a reality.

Nevertheless, only two years ago we predicted that it would take another ten years to reach the point we have reached today.”