Network‐based analysis predicts interacting genetic modifiers from a meta‐mapping study of spike–wave discharge in mice

Absence seizures are characterized by brief lapses in awareness accompanied by a hallmark spike‐and‐wave discharge (SWD) electroencephalographic pattern and are common to genetic generalized epilepsies (GGEs). While numerous genes have been associated with increased risk, including some Mendelian forms with a single causal allele, most cases of GGE are idiopathic and there are many unknown genetic modifiers of GGE influencing risk and severity. In a previous meta‐mapping study, crosses between transgenic C57BL/6 and C3HeB/FeJ strains, each carrying one of three SWD‐causing mutations (Gabrg2tm1Spet(R43Q), Scn8a8j or Gria4spkw1), demonstrated an antagonistic epistatic interaction between loci on mouse chromosomes 2 and 7 influencing SWD. These results implicate universal modifiers in the B6 background that mitigate SWD severity through a common pathway, independent of the causal mutation. In this study, we prioritized candidate modifiers in these interacting loci. Our approach integrated human genome‐wide association results with gene interaction networks and mouse brain gene expression to prioritize candidate genes and pathways driving variation in SWD outcomes. We considered candidate genes that are functionally associated with human GGE risk genes and genes with evidence for coding or non‐coding allele effects between the B6 and C3H backgrounds. Our analyses output a summary ranking of gene pairs, one gene from each locus, as candidates for explaining the epistatic interaction. Our top‐ranking gene pairs implicate microtubule function, cytoskeletal stability and cell cycle regulation as novel hypotheses about the source of SWD variation across strain backgrounds, which could clarify underlying mechanisms driving differences in GGE severity in humans. Absence seizures are characterized by brief lapses in awareness accompanied by a hallmark spike‐and‐wave discharge (SWD). In a previous mapping study, crosses between transgenic C57BL/6 and C3HeB/FeJ strains, each carrying one of three SWD‐causing mutations demonstrated an antagonistic epistatic interaction between loci on mouse chromosomes 2 and 7 influencing SWD. In this study, we computationally prioritized candidate modifiers in these interacting loci, implicating microtubule function, cytoskeletal stability and cell cycle regulation as novel hypotheses about the source of SWD variation in mice.