Experimental analysis of background dependent effects of genetic interactions can be designed using strains generated by introgression of small genetic regions containing identical genotypes at loci in question into different inbred strains. We use a novel multilocus paradigm, denoted conditional intergenic functional association (CIFA), to simulate this procedure, with the trade-off of power for convenience that is affordable when sufficiently strong effects are present. We analyze nine enzyme loci at three chromosomes in groups of D. melanogaster with different developmental rates that showed similar allelic frequencies at individual loci. Results obtained suggest the presence of adaptive interaction between particular alleles at two loci when genetic variation at seven background loci is eliminated. Biochemical considerations show that, in the resulting developmental context, strong interaction between these genes may emerge from shifted control of the pentose phosphate pathway, with cascading effects on the glycolysis, TCA cycle, and biosynthetic pathways: one gene may assume control of the irreversible rate-limiting step in the pentose phosphate pathway, whereas the other gene may assume control of the NADP+ level that regulates the same rate-limiting step as an electron acceptor. The newly developing functional genomics research and the absence of inbreeding make CIFA directly applicable to complex human traits in large samples.