Some short conclusions and perspectives

This section is intended to show the problems that may arise when studying very flexible gas phase models.

We will not comment the differences between QM/MM results in section 2.3 and the results obtained here. Both results are not comparable because we cannot ensure the connection of the intermediates under a unique MEP. The transition states found in the previous QM/MM model are also found here, but in general we can hardly connect the TS and the several intermediates due to the tricky behavior of the optimizations. Different structures have been found and all of them are possible candidates to the same intermediate.

In the QM/MM study of Mandelate Racemase we also found different minima in small energetic range, but, as we already pointed out, all the minima had the same structure in the active site. This is not the case in these gas phase calculations. There are some interactions that we did not find in the QM/MM model and we attribute it to the intrinsic problem of constriction of some coordinates.

Probably the problems encountered in the work of Alagona et. al.[232] and already commented in page were due to these same problems, the difficult combination of the needed constraints and the flexible active site of Mandelate Racemase.

This is why we think that, in this case, QM/MM methods are not only an improvement to the solute/solvent interaction but an easier way to obtain reliable structures. For these reasons we insist that the exploration of big-dimensioned QM/MM surfaces with optimization methods is a valuable tool. In the next chapter we will present a method to carry out such exploration.