One more thing:Is it possible to retain point group symmetry during geometry optimization for molecular systems in CP2K?