Figure 5

Fingerprinting of the yeast regulatory activity network using Quontology (H.J.B. and R.L., unpublished work, see Materials and methods). Functional categories are induced (red) or repressed (green) by overexpression of HAP4 or in the diauxic shift [1]. The whole-genome expression data of the diauxic shift are divided into three datasets that represent: glucose excess; a strongly decreasing glucose concentration; and glucose depletion. Using the quontology algorithm, Z-scores were calculated for the functional categories defined by MIPS [8]. Generally, all categories that have at least one Z-score larger than 4 or smaller than -4 are listed (p < 0.1). The hierarchy of the categories is preserved: the main categories in capitals, the subcategories indented and in lower case. A filled circle marks MIPS categories that show the same pattern of expression in more respiratory conditions (the five columns on the right). Interestingly, overexpression of HAP4 results in specific induction of MIPS categories that are involved in mitochondrial function, causing the regulatory activity network to move to a state that is different from the diauxic shift.