FlyAtlas: a comprehensive online atlas of gene expression in Drosophila

Multicellular organisms, from humans to flies to plants, have distinct tissues specialized for distinct tasks. What makes one tissue different from another? Every cell in the body contains the same genetic code (of 15000-25000 genes, depending on the organism); but different tissues switch on (express) different subsets of these genes. For example, skin is skin because it expresses keratin, and liver is liver because it expresses detoxifying enzymes. To understand how organisms work, therefore, we need to know which genes are expressed in which tissues. There are many techniques for mapping gene expression levels, but most are laborious and prone to error. By contrast, DNA microarrays allow the relatively precise measurement of expression levels of all genes in a genome in a SINGLE experiment. A comprehensive, microarray-based 'atlas' of gene expression in specific tissues and cell types thoughout an organism is thus a fantastically useful resource to a whole community; if the job is done properly, there is no need to do it again. This proposal is to extend such a service for the Drosophila (fruit fly) research community, both in the UK and worldwide. A pilot project (with 9 adult tissues) has attracted over two thousand users in its first 6 months, confirming the need for such a resource. We propose to extend the resource to further tissues and further life stages, with the data easily accessible on the web (at flyatlas.org). The resource will allow scientists to target their effort more precisely, saving money and time, and reducing unnecessary animal use.

The Drosophila research community is large, fast-moving, and has contributed greatly to our understanding of organismal function, from development to behaviour. With the publication of the Drosophila genome project, it has become clear that, for at least half the 13500 or so genes, there is no known function. A reverse genetic strategy in a simple organism like Drosophila is clearly a potent tool for elucidating gene function - but what phenotypes should be studied? Clearly, the answer depends on the tissues in which a gene is expressed. Rather than leave post-docs worldwide labouring with Northerns, RT-PCRs or in situs of variable quality and comprehensiveness, it would make sense to do the job once, properly. Flyatlas (flyatlas.org) puts experimenters in touch with their tissues quickly, clearly and unambiguously. The original dataset, generated through BBSRC IGF funding, comprises 9 adult tissues, two larval tissues, and a whole-fly sample. It is based on Affymetrix Dros2 chips, with 18770 probesets to cover 13500 genes, and so is accurate and comprehensive. Since its launch in the autumn of 2006, it has attracted over a thousand users, confirming the hunger of the community for such data. This proposal is to extend the dataset to further adult tissues and further life stages, to provide a one-stop lookup-shop for Drosophila expression, and a rich bioinformatic resource for novel meta-analysis.