eXPRESSION: an in silico tool to predict patterns of gene expression

Abstract

In embryological studies, expression pattern analyses are of special importance since genes that have temporally and spatially restricted expression are not only essential as lineage markers but are often causative in formation of specific fates. Further, where a molecule is expressed can be quite revealing in regard to its endogenous function. We present a gene discovery tool, termed eXPRESSION, that utilizes the public EST databases to identify genes matching desired transcriptional profiles. We first tested and validated the ability of eXPRESSION to discover tissue-specific genes in the adult mouse; empirically as well as with DNA microarrays and RT-PCRs. These studies showed that eXPRESSION predictions could identify genes that are specifically expressed in adult mouse tissues. Next, we developed a novel search strategy to find genes that are expressed in specific regions or tissues of the developing mouse embryo. With these tools, we identified several novel genes that exhibited a neural-specific or neural-enriched expression pattern during murine development. The data show that eXPRESSION is widely applicable and may be used to identify both adult and embryonic tissue- or organ-specific genes with minimal cost and effort.