Gene expression data generated using microarrays can be used to group genes into clusters that reflect their involvement in biological processes1. We have more than tripled the gene expression data available to include approximately 25 physiologically different conditions assayed with approximately 400 individual array hybridizations (each hybridization measures mRNA levels for each gene in the entire genome). This data set now contains characterizations of gene expression changes in response to numerous stresses, as well as changes during the cell cycle, sporulation and nutritional conditions. Such a large data set (∼2.5 million independent observations) permits only a cursory description in words; instead, these data are the foundation for a reference physiological gene expression database. An example of the kinds of information one can extract from these data is the genome-wide pattern of gene expression in response to stress. In addition to generic negative responses (for example, turning down transcription of the genes responsible for translation or progress through the cell cycle), there are a large number of genes that are induced by virtually any stress. There are far fewer genes whose induction is seen only in a particular type of stress. Of particular interest is the overlap in gene expression between oxidative stress and heat shock. We hope that this data set will be useful in providing a context for a great variety of physiological measurements in the future.
References
Eisen, M.B. et al. Proc. Natl Acad. Sci. USA 95, 14863–14868 (1998).
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Spellman, P., Gasch, A., Eisen, M. et al. Functional clustering of genes using microarray gene expression data. Nat Genet 23 (Suppl 3), 75 (1999). https://doi.org/10.1038/14406
Issue Date:
DOI: https://doi.org/10.1038/14406