Databases of DNA sequences, physical maps, genetic maps, gene polymorphisms, protein structures, and gene expression have produced a need for systematic quantitative analysis that goes by the name of bioinformatics. To manage these large datasets in the domain of functional genomics requires algorithmic implementation on computers. Functional genomics will continue to be hypothesis-driven and hypothesis-generating biological research. Because the datasets are of high dimensionality, involving relatively small numbers of cases, a large number of solutions can explain the data; the computational techniques to unravel these involve both supervised and unsupervised learning.