Natural selection acts on variation that occurs by spontaneous mutation, and the fate of mutations depends on their selective value and drift. Moreover, mutations are distributed within and amongst populations due to demographic events, and often reflect historical biogeography and migration. The evolution of traits by natural selection in local populations depends on the rate of mutational input and draws upon standing variation. Quantifying the relative importance of both is of interest especially in complex traits encoded by interacting genes where major adaptive mutations may require allelic reconfiguration at interacting loci. Here I address the role of complex demographics and standing variations in assembly of a polygenic adaptive trait – warfarin resistance in natural populations of Norway rat (Rattus norvegicus). Warfarin is used both as a rodenticide and a prescribed blood thinning medicine in humans. It works by inhibiting the vitamin K cycle, specifically blocking Vkorc1 (vitamin K epoxide reductase complex subunit 1). Warfarin resistance mainly encoded by Y139C mutation in the Vkorc1 evolved quickly in wild Norway rats. In humans, warfarin has a narrow therapeutic window and depends upon variation in at least 2 additional genes suggesting a complex genetic architecture. In this dissertation, I show that association study and population genetics of any variant in the study area would need to consider the population structure at local and regional scales. A major contribution of this dissertation is a fully sequenced, mapped and annotated genome sequence of out-group species, the roof-rat (Rattus rattus) which enables the polarization of variants in the Norway rat as either ancestral or derived. This work also showed that the causative variation in Vkorc1 is a new mutation and such instances of new mutations are rare in the genome of Norway rats. Interestingly, I found that the majority of other genes participating in Vitamin K cycle carry common standing variants whose local occurrences may have been affected by demographics more strongly than by selection. Selection on interacting putative candidate genes underlying warfarin resistance in rats may have favored a main de novo mutation in Vkorc1, and previously rare standing derived variants in interacting genes.