AbstractThe phenotype of an individual can be affected not only by the individual’s own genotypes (direct genetic effects, DGE) but also by genotypes of other, interacting individuals (indirect genetic effects, IGE). IGE have been detected, using polygenic models, for a broad range of biomedical phenotypes. However, little is known about the loci, genes and traits of interacting partners mediating the effects, especially where non-familial IGE are concerned. To address this question, we studied IGE arising in cages of unrelated, adult laboratory mice. We leveraged a dataset of 170 behavioural, physiological and morphological phenotypes measured in 1,812 genetically heterogeneous mice and developed two approaches. First, we used variance components models to estimate, for each phenotype, the correlation between DGE and IGE. Our results demonstrate some overlap but also differences between the mechanisms of DGE and IGE for a given phenotype. Second, we developed and applied methods for the genome-wide association study of IGE (igeGWAS) in order to identify loci, genes and traits of cage mates mediating IGE. We identified 24 genomewide significant IGE loci for 17 phenotypes (FDR < 10%), none of which overlapped with genome-wide significant DGE loci for the same phenotype. Using an exhaustive list of single nucleotide polymorphisms at each locus, we fine-mapped each association. At six loci this pointed to a single candidate gene, which we used to formulate specific hypotheses as to the mechanisms of IGE. The empirical insights from our study and the analytical strategies we have developed pave the way for using igeGWAS to unravel mechanisms of phenotypic variation that are expressed only in the context of social interactions.