Computers are now so central to genetics — and most other branches of biology — that computational expertise is an essential part of any research programme. The rise of computational biology has been driven in part by the acquisition and accumulation of data mountains, and learning how to explore these mountains is making possible entirely new types of experiment. This issue of Nature Reviews Genetics features a Focus, covering three very different ways in which computational analysis is being used in genetic research.

Duncan Davidson and Richard Baldock discuss bioinformatic frameworks, which allow information, such as expression or functional data, to be compared in a rigorous and reproducible way. A CD-ROM version of the Edinburgh Mouse Atlas — a framework for mouse development — is also provided with this issue. John Quackenbush presents a survey of the computational tools that are used for microarray analysis. These tools are generally less well understood than the laboratory protocols for microarray studies, but choices concerning data processing and analysis can have a profound influence on the interpretation of the experiment. Finally, James Foster discusses how our knowledge of evolutionary genetics is being used to devise computational tools themselves. Computational evolution is a field that uses the principles of evolution to build new programs and even hardware. One fascinating by-product is that the nature of the tools that evolve mimics some of the features of evolved genetic systems — robustness being a most striking example.

Besides the Focus, this issue also sees the first in a collection of articles that examines the origins of different model systems. This month, Rachel Ankeny writes about Caenorhabditis elegans. Computational biology can be a futuristic discipline, so it's good to stay in touch with the roots of our subject too.