I am trying to perform an ab initio molecular dynamics over a system in a potential energy well. My concern is about how to initialize the velocities of atoms using the Maxwell-Boltzmann velocity distribution to a finite temperature. The idea is to convert the potential energy stored in the system to the kinetic energy that will be shared among the degrees of freedom, therefore obtaining the Maxwell-Boltzmann distribution for an equivalent temperature, but I'm not sure it's the correct procedure and if I'm overlooking something