AbstractRNA-seq is commonly used to identify genetic modules that respond to perturbations. In single cells, transcriptomes have been used as phenotypes, but this concept has not been applied to whole-organism RNA-seq. Linear models can quantify expression effects of individual mutants and identify epistatic effects in double mutants. To make interpretation of these high-dimensional measurements intuitive, we developed a single coefficient to quantify transcriptome-wide epistasis that accurately reflects the underlying interactions. To demonstrate our approach, we sequenced four single and two double mutants ofCaenorhabditis elegans. From these mutants, we reconstructed the known hypoxia pathway. In addition, we uncovered a class of 56 genes that have opposing changes in expression inegl-9(lf)compared tovhl-1(lf)but theegl-9(lf); vhl-1(lf)mutant has the same phenotype asegl-9(lf). This class violates the classical model of HIF-1 regulation, but can be explained by postulating a role of hydroxylated HIF-1 in transcriptional control.Significance StatementTranscriptome profiling is a way to quickly and quantitatively measure gene expression level. Because of their quantitative nature, there is widespread interest in using transcriptomic profiles as a phenotype for genetic analysis. However, a source of major concern is that whole-animal transcriptomic profiles mix the expression signatures of multiple cellular states, making it hard to accurately reconstruct genetic interactions. Additionally, it has been difficult to quantify epistasis, the signature of genetic interaction between two genes, in these molecular phenotypes. Here, we show that it is possible to accurately reconstruct genetic interactions between genes using whole-animal RNA sequencing, and we demonstrate a powerful new way to measure and understand epistasis arising from these measurements. This suggests that whole-organism RNA-seq can be a powerful tool with which to understand genetic interactions in entire organisms and not only in isolated cells. With the advent of genome engineering tools, generating mutants has become easier and faster for many organisms. As mutants become easier to create, phenotyping them has become a major bottleneck in understanding the biological functions of the genes in question. Our work presents a possible solution to this problem, because transcriptome profiling is fast and sensitive to genetic perturbations regardless of the context they operate in.