It is no longer news that the number of scientists trained to translate basic research into clinical practice has been declining over the past two decades. And that, at the same time, the huge expansion in genomic and proteomic data has meant that the quantity of things to translate has increased enormously. The career problem, crudely stated, is there are too many basic biologists and not enough physician scientists.

Participants at a 'Days of Molecular Medicine' conference this March agreed that one problem might be that the traditional MD/PhD pathway is no longer the best way to train scientists who want to translate findings from the bench to the bedside. Such programmes take too long, cost too much, and offer too few slots to meet the existing need.

The answer, the conference participants suggested, is to find different routes and incentives. Nature Medicine, which organized the meeting with the University of California, San Diego, and the Salk Institute, this month published three solutions that the conference came up with (full text is on the Naturejobs Life Sciences Channel).

One is to close the gap from the 'other side', by giving some PhDs more medical training, even allowing them to spend time on rounds with medical residents (see Nature Med. 8, 433–436; 2002). Another solution is to find ways to encourage more 'late bloomers' — medical students who decide midway through their training that they have an interest in research (see Nature Med. 8, 437–439; 2002). Finally, as women are disproportionately under-represented in the MD/PhD ranks, policy-makers should look at ways to reverse this trend (see Nature Med. 8, 439–441; 2002).

Although no one approach is likely to be a panacea, combining several might help biologists to find jobs and turn data into drugs.