Aiming at multi-objective and priority-adjustable heat transfer surface optimization in turbulent flows, a new adjoint-based shape optimization algorithm with DNS(direct numerical simulation)-informed RANS approach is developed. The cost functional for the optimization is set as a linear combination of Stanton number and Fanning friction factor with the weight coefficient ω ∈ [0,1] to determine the priority between the heat transfer enhancement and the pressure loss reduction corresponding to ω = 0 to 1, respectively. In this study, the dependency of ω on the resultant optimal shape is systematically investigated. The algorithm is examined by applying to a V-shaped oblique wavy fin between two parallel walls. The constant bulk flow rate and uniform heating of the fluid is assumed. The mechanism of improved performances are analyzed for the typical cases, i.e. ω = 0.0,0.5,1.0. It was found that the characteristic structure in each resultant shapes such as notches, protrudes, holes and flatted baseline of wave of the fin affect on the flow as following: (1) with ω = 0.0, the heat transfer is enhanced by large swirling flow structures generated by the finger-like protrusion and holes them, (2) with ω = 0.5, the pressure loss is reduced and heat transfer is sustained by flattened wavy shape and notches near side wall due to the edge effect, and (3) with ω = 1.0, the pressure loss is drastically reduced by the large notches near the side wall widening the streamwise cross-sectional area.