Abstract
The determination of the human genome sequence has spurred a great deal of interest in using changes in levels of gene expression in individuals to discover the basis of disease and to identify new drug targets. While this process has been successful and there are indeed specific gene changes in certain diseases, it has also recently been shown that variation in life span and longevity in identical and nonidentical twins is mainly explained by environmental effects such as smoking and diet [1]. This new approach has been incorporated into drug discovery programs in pharmaceutical companies and although some significant advances have been made-notably in some aspects of understanding cancer susceptibility [2,3]—it often remains difficult to relate any changes seen to conventional end points used in disease diagnosis and to optimize efficacy and minimize toxicity in pharmaceutical development. For example, in toxicogenomics studies, candidate drugs can give rise to many gene expression changes that have no actual pathological consequences.