We have a growing capacity to test for genes that cause or predispose to disease, but the take-up of this technology is not straightforward.

Some of our genes affect our chances of contracting diseases. So the idea that it would be good to know the details of the genetic hand you have been dealt has been around for many years. But as it has come nearer to reality, the application has become less straightforward.

On one hand, tests may tell you things you can do nothing about, like identifying who in a family with a history of Huntington's disease will definitely develop the condition. Many people offered this test decide they would rather not know.

On the other hand, tests may be a very poor guide to what might actually happen, so although you may be genetically more likely to get a particular disease, this does not necessarily mean you will go on to develop it. Most diseases are affected by lots of different things, and genes play a small part. The gene tests now marketed over the internet generally fall into this category. They are unlikely to provide much useful information.

There are also problems with making the tests reliable and explaining to people how they should be interpreted (usually carried out by genetic counsellors). The gene which, if both copies have an altered form, causes cystic fibrosis (CF) can have over 500 different mutations. Tests only pick up the more common ones. Then there is the practical issue of getting a technology adopted by the NHS. CF screening is offered in some but not all areas.

So even with this relatively simple biological situation, the introduction of new technological approaches is hugely complex.

And that's without public fears about insurance companies using genetic tests to refuse people health cover, life insurance or access to mortgages…