A computational analysis of whole-genome expression data reveals chromosomal domains of gene expression

Chromosome correlation maps display correlations between the expression pat terns of genes on the same chromosome. Using these maps, we show here that adjacent pairs of genes, as well as nearby non-adjacent pairs of genes, sho w correlated expression independent of their orientation. We present specif ic examples of adjacent pairs with highly correlated expression patterns, i n which the promoter of only one of the two genes contains an upstream acti vating sequence (UAS) known to be associated with that expression pattern. Finally, we show that genes with similar functions tend to occur in adjacen t positions along the chromosomes. Our results suggest that, in certain chr omosomal expression domains, an UAS can affect the transcription of genes t hat are not immediately downstream from it.