Quantitative genetic analysis relies on measuring phenotypes of single and double mutants. If two genes interact, the double mutant will have a phenotype that is not the sum of the phenotypes of the single mutants that were combined to make up its genotype. Although epistatic analysis lies at the heart of genetic analysis, its generalization to multi-dimensional quantitative phenotypes has proven very difficult. Transcriptional profiles are a particularly appealing complex quantitative phenotype, given the ease of obtaining RNA and the accuracy with which RNA levels can be measured. Despite its quantitative resolution of transcript levels, RNA-seq is used mostly as a qualitative tool with which to identify genes that are responsive to a perturbation relative to a control. Recently, we have developed tools that enable us to consider transcriptomes as phenotypes by comparing how gene expression levels change across two single mutants and comparing them with the expression levels in a double mutant. We can summarize these results in a single number, called the transcriptome-wide epistasis coefficient, which accurately reflects the underlying genetic architecture. As a case study, we chose genes associated with the synthetic Multivulva (synMuv) phenotype. Briefly, synMuv genes can be separated into two functionally redundant classes named A and B. Class B largely consists of general chromatin modifiers, such as
epc-1 enhancer of polycomb,
hpl-2 heterochromatin protein 1, and
lin-35 Rb. Class A is composed of nematode-specific genes with poorly understood molecular functions, but that act redundantly with the class B synMuv genes to prevent ectopic expression of
lin-3 EGF. In synMuv AB double mutants, ectopic expression of
lin-3 induces ectopic vulval development, the synMuv phenotype. How or why this happens is not well understood despite work on the synMuv genes since the 1980s. We have sequenced several synMuv class A and class B single mutants and AA and AB double mutants to understand how these two redundant pathways interact in a whole-organism framework.