Summary/Abstract A broad goal for biologists is to understand how genetic variation influences phenotype, including development, physiology, and disease. For complex behaviors, it remains difficult to trace a path from the genome to particular cell populations to brain function. In the proposed research, we will identify how natural variation in vertebrate social behavior is encoded within the genome and executed in the brain via context dependent gene expression in specific cell types. To accomplish this, we will utilize the assemblage of cichlids from Lake Malawi, East Africa to take advantage of recent evolutionary radiation that generated large differences in social behavior with minimal genetic divergence. We will use new automated methods to quantify differences in male bower building behavior. We will map forebrain cell types recruited during bower building and identify cis-regulatory circuits in these brain regions, using RNA-seq and ATAC-seq from single cells. Finally, we will use QTL mapping and the construction of near-isogenic lines (NILs) to identify the genetic basis of bower behavior and link genome regions to specific behavioral and cellular subroutines. This work will identify cell populations, gene expression programs and neuronal circuits regulating social behaviors in outbred vertebrates and provide comparative insights into human behavioral diversity including common neurological diseases.