A curriculum for bioinformatics: the time is ripe

There seems to be wide agreement within both industry and academia that there are not enough scientists adequately trained in bioinformatics or computational biology. This sentiment stems principally from the difficulties in finding employees, graduate students and post-docs with appropriate skills for joining research and/or development teams in this field. The recent drain of academics into industry threatens to reduce our ability to provide the training needed to meet the demand of the job markets. An obvious question is 'What is the proper curriculum for bioinformatics professionals?' At first, the idea of defining a curriculum for bioinformatics may seem premature. The very definition of bioinformatics is still the matter of some debate. Although some interpret it narrowly as the information science techniques needed to support genome analysis, many have begun to use it synonymously with 'computational molecular biology' or even all of 'computational biology'. For this discussion of curriculum, bioinformatics addresses problems related to the storage, retrieval and analysis of information about biological structure, sequence and function. There are currently two models for training. In the first model, post-doctoral fellows with core training in a technical field (such as computer science) or in a subdiscipline of biology receive speciality training in computational biology in order to become a 'computer scientist who specializes in biology' or a 'biologist who specializes in computer science'. While a valuable strategy, the post-doctoral model suffers because it is an expensive way (both in time and resources) to train individuals — learning the 'other' field is in many cases like going back to graduate school. In the second model, therefore, graduate students are trained primarily in bioinformatics or computational biology, without a preliminary training in one of the contributing disciplines. The curriculum for these students must provide them with a skill set that is long-lived and endures beyond the current fads of what is considered 'hot'. These students will not have traditional biological science or technical training to fall back on, and so it is critical that we provide the next generation with skills to solve industrial and academic problems that we cannot anticipate. I approach the problem of defining curriculum with a bias that bioinformatics is not simply a proper subset of biology or computer science, but has a growing and independent base of scientific tenets that requires specific training not appropriate for either biology or computer science alone. An appropriate academic curriculum for the …