In this talk we show a simple formulation for the doubly nonnegative (DNN) relaxation of the protein side-chain positioning (SCP) problem. We take advantage of the natural splitting of variables that stems from the facial reduction technique in the semidefinite relaxation. We solve the DNN relaxation using a variation of the Peaceman-Rachford splitting method. We obtain strong results from our numerical experiments; we solve all our instances of the SCP problem close to optimality.