Interpretation of the GWAS results was facilitated using network reconstruction and visualization. A network was created using SNPs in windows that explain the highest proportion of the variance for a particular trait. Selected SNPs were located in the rainbow trout genome (Berthelot et al., 2014) using the Golden Helix Genome Browser v2.1.0 (Golden Helix Inc.). When a SNP was mapped to a gene (exon or intron), the nucleotide sequence was blasted to the Mus musculus genome using NCBI tools to predict the orthologous genes. The mouse genome was selected because it is well annotated, is available in several bioinformatics tools, and has been used as a model organism for many years (Guenet, 2005). With the list of the genes, a network was visualized with Cytoscape (Shannon et al., 2003; Serão et al., 2013) using the BisoGenet plug-in (Martin et al., 2010; Gonzalez-Pena et al., 2016). The BisoGenet plug-in uses empirical and predicted DNA-DNA, DNA-protein, and protein-protein interactions to reconstruct and visualize likely mouse gene networks. Only gene-gene interactions were represented and only one neighbor connecting the genes was admitted. The edges denote known relationships between genes from several databases summarized in the SysBiomics repository that integrates data from NCBI, UniProt, KEGG, and GO databases using the k-nearest neighbor model. The node color denotes if the gene was identified (green) directly from the GWAS or if it is a connecting neighbor (pink) based on annotation of the mouse genome.