Fig. 5
Buffering by induced dependency. a Genetic interaction between Hac1 and Rpn4. Representation as described for Fig. 4a. Grey depicts non-epistatic expression changes. Available DNA binding data showing significant overlap with at least one gene set are presented below the expression data. These are in vivo binding targets (“1-in vivo” as in [26]) as well as in vitro-derived promoter affinity scores (“2-in vitro”,“3-in vitro” as calculated from [28] and [27], respectively; Methods). Promoter affinity scores range from zero (white) to three (black) as depicted. Significant correspondence with expression data is depicted by red boxes. Top GO-BP terms are: gene set 2: protein refolding (P = 2.85 × 10−6); gene set 3: modification-dependent catabolic process (P = 1.29 × 10−32); gene set 4: transmembrane transport (P = 0.04); gene set 5: de novo pyrimidine nucleobase biosynthetic process (P = 9.79 × 10−3); gene set 6: response to water deprivation (P = 0.03); gene set 7: glutamine family amino acid biosynthetic process (P = 2.7 × 10−4); gene set 8: ribosome biogenesis (P = 1.55 × 10−21); gene set 9: oxidation-reduction process (P = 1.1 × 10−6). b Rpn4-activated genes (zoom-up of gene set 3 in a). Red labels for annotated target genes of Rpn4 in vivo [26]. c Hac1-activated genes (zoom-up of gene sets 1 and 5 in a). Red labels for annotated target genes of Hac1 in vivo [26]. d Cartoon depicting the proposed genetic interaction between Hac1 and Rpn4. Consequences of individual deletions are indicated in red. e Summarized model to describe the proposed genetic interaction between Hac1 and Rpn4. f Generalized model for “buffering by induced dependency”