An overview of the major first-principles methods used to simulate condensed phases is presented, with special emphasis on chalcogenide glasses. The scope of this review article is to offer a survey of fundamental algorithms and techniques, accompanied by a few recent examples particularly representative of computational materials science applied to disordered chalcogenide phases. Special attention is devoted to the inclusion of long-range van der Waals dispersion forces, treatment of the exact exchange, dynamical simulations and extraction of optical and dielectric properties. Machine learning techniques are introduced as recent forefront applications of first-principle methods. In this latter case, accurate quantum-mechanics based simulations are crucial to generate a data base exploited by neuronal-network type algorithms to create accurate interatomic potentials (force fields) allowing for large and long-lasting simulations of realistic disordered materials. The atomic-level knowledge provided by the combination of high-performance computing and advanced computational methods pave the route for a rational approach to the design of novel chalcogenides possessing tuned properties for specific applications in next-generation devices.