This protocol minimizes the molecule (to relax any "hot spots" in the molecule), and heats the molecule to a set temperature. The molecule is then "cooked" at this temperature for a long time; the molecular conformation is saved at regular intervals. The ensemble of conformations represents the possible conformations at the set temperature, assuming that the dynamics is performed long enough to sample the entire range of conformations. Note also that the ensemble represents the POSSIBLE conformations; an abundance of one type of conformation does not necessarily indicate that this conformation is the most PROBABLE.
This protocol can be used to determine which parts of the molecule are most dynamic (according to the molecular modeling forcefield).
This protocol first uses the Molecular Dynamics protocol. After the molecular dynamics protocol is finished, each conformation is recalled and "quenched": the conformation is minimized to the nearest local minimum. The ensemble of conformations represents the possible energy minima, assuming that the dynamics is performed long enough to sample the entire range of conformations.
Note also that the ensemble represents the POSSIBLE energy minima; an abundance of one type of conformation does not necessarily indicate that this energy minimum is the most PROBABLE. However, assuming that the dynamics is performed long enough to sample the entire range of conformations, the population of each conformation is roughly weighted by the volume of it's global energy well. For example, the following diagram represents a single term of the force field. Starting at Conformation C, simulated annealing is performed at a temperature of 900K. After the dynamics and simulated annealing, Conformations A, B, and C are present; because of the heights of the energy barriers, conformation D is not sampled at 900K. The relative population of conformer A, B, and C roughly corresponds to the areas of their potential energy wells.
For molecules with only a few energy minima (possibly due to fixing or constraining much of the molecule), this method works well. However, most molecules have many local energy minima, so this method results in many different conformations.
This protocol minimizes the molecule (to relax any "hot spots" in the molecule), heats the molecule to a set temperature, then slowly cools the molecule, and finally minimizes the molecule. This procedure is repeated again and again to generate a large ensemble of conformations. The ensemble of conformations represents the possible energy minima, assuming that the dynamics is performed long enough to sample the entire range of conformations.
The ensemble represents the POSSIBLE global energy minima; an abundance of one type of conformation does not necessarily indicate that this energy minimum is the most PROBABLE. However, assuming that the dynamics is performed long enough to sample the entire range of conformations, the population of each conformation is roughly weighted by the volume of it's global energy well AT THE FINAL TEMPERATURE OF THE SLOW COOLING, BEFORE THE QUENCHING. For example, the following diagram represents a single term of the force field. Starting at Conformation C, simulated annealing is performed at a temperature of 900K. After the dynamics and slowly-quenched simulated annealing (slowly cooled to 300K), Conformations A, B, and C are present; because of the heights of the energy barriers, conformation D is not sampled at 900K. The relative population of conformer A, B, and C roughly corresponds to the areas of their potential energy wells relative to 300K.
The simulated annealing protocol described above calculates a new conformation starting from a fully minimized structure---the structure is heated and cooled, heated and cooled, heated and cooled.....If your potential energy surface has a very large potential energy well with a wide "top" at the selected dynamics temperature, then this procedure may cause the simulated annealing to spend a lot of time in this large well, since each time the structure must "climb out" of the potential energy well from the very bottom of the well.
A variation of the simulated annealing protocol can be used to more quickly and easily search conformation space. This protocol minimizes the molecule (to relax any "hot spots" in the molecule), heats the molecule to a set temperature, cools the molecule, heats the molecule, cools the molecule, etc,., but DOES NOT minimize the molecule during the simulated annealing. After the slowly-quenched simulated annealing protocol is finished, each conformation is recalled and "absolutely quenched": the conformation is finally minimized to the nearest local minimum.
The ensemble represents the POSSIBLE energy minimia, assuming that the dynamics is performed long enough to sample the entire range of conformations. An abundance of one type of conformation does not necessarily indicate that this energy minimum is the most PROBABLE. The population of each conformer is roughly weighted by the area of the potential energy well, as described above.