Classes in computational biology are among the most popular offerings of the advanced-course season. But they aren't the only show in town. For example, one of Cold Spring Harbor Laboratory's stalwarts, a class on mouse embryology, celebrated its twentieth anniversary this month. And it still draws four or five times as many applicants as it can accommodate, even though — or perhaps because — “the course in no way bears any resemblance to what was taught 20 years ago”, says Dave Stewart, director of meetings and courses at the laboratory.

Change, when it comes to course curricula, does scientists good. And there definitely is change afoot, with computational components taking up larger portions of the curriculum for subjects such as proteomics, genetics and gene expression. This allows more discipline-specific techniques to be disseminated — such as methods of interpreting microarray data, or using computers to analyse different kinds of neuroscience imaging data. Of course, these tailored courses complement the more general bioinformatics offerings.

Interest in both tailored and general bioinformatics training shows no sign of waning. A bioinformatics course sponsored by the French research agency INSERM is already oversubscribed with just under a month to go before the closing date for applications. In addition, almost all of the UK Wellcome Trust's genomics and proteomics courses feature computational components, and informatics is a growing aspect in many of Cold Spring Harbor Laboratory's courses — but not, of course, the mouse-embryology one.

So how can the enduring appeal of the mouse course be explained in these high-tech times? Stewart believes he knows: “A lot of people are still interested in learning from regular biology.”