It is well known that a complex disease stems from the malfunctions of some biomolecular networks which control the pathogenesis of the complex disease. Disease genes can be viewed as special nodes in biomolecular networks and are not randomly distributed (located) from existing evidences. The identification of disease genes is critical towards the understanding of complex diseases. We formulated identifying disease genes from a biomolecular network as finding the most possible configuration of the network. Borrowing from statistical mechanics, we employed the Boltzmann distribution to compute the probability of configurations of a network while adopting the Ising model to define the network energy. In this talk, I will present some of our recent results for identifying disease genes from biomolecular networks based on our proposed statistical mechanics-based methods.