Those of us who dedicate ourselves to Chemistry in any of its variants, owe a debt of gratitude to Karplus, Levitt and Warshel. They, like others over the years, have focused their research on the development of programs that allow us to develop models that simulate reality Chemistry, by means of which we can better plan, analyze and develop our work. And I believe that the debt is even greater for those of us who dedicate ourselves to design and drug synthesis. The long and costly journey of the new drug has been significantly shortened by the application of these programs, since we now have models that simulate the chemical and biological systems involved in the computer.

As an example, let us imagine that we want to produce a drug. Using starting products, reagents, and applying a synthesis method, we are going to obtain a new compound or reaction product in which we can recognize parts of the starting products joined together in another way, a new molecular entity. During the reaction, some chemical bonds are broken and others are created, forming the products of the reaction; there are steps that sometimes only last a fraction of a millisecond, intermediate chemical species are generated, intermediate states, generally excited, not visible, unstable, which must exist if the reaction takes place. Electrons pass from one point to another, there are orbitals that are emptied and others are filled, some molecules are "destroyed", others are "created".

In order to visualize these steps and control them and correct the method, if necessary, we now have models that simulate them and, by studying them, we obtain data that help to optimize them. There are programs that allow the generation of 3D models of reagents, products and intermediate species, and of biomolecules, computer heirs of those models of wooden or plastic spheres and rods, which allow the molecules to be "seen" on the computer and studied.

The study can be done by applying classical Newtonian mechanics, treating the atoms as if they were spheres, with specific mass and volume, joined together by springs, bonds with a specific force. Electrons are not explicitly considered, but implicitly, so they serve to obtain a very approximate idea of what the molecule is like, what preferred form it adopts at rest, its conformation or conformations, but they do not allow us to know what has happened to the electrons, the charges, those "responsible" for the reaction. We are at an atomic level where the classical approaches are still valid. The speed and possibility of applying them to large molecules motivate their employment.

To find out what has happened in the reaction, where the electrons have gone, how the bonds have been broken and others formed, what reaction intermediates appear, etc., we must turn to quantum physics, since electrons are sub-atomic particles. In this way we obtain data which allows us to explain reaction mechanisms, optimize procedures, the need to use catalysts, etc. But there is a problem: these calculations are complicated in time and power, and only applicable to small molecules. For this reason, these are analyzed by both classical and quantum methods, depending on the data of interest, while only classical methods can be applied to large molecules. The data obtained are useful and are routinely used, but they do not provide correct information about the system.

Thanks to work , developed by Karplus, Levitt and Warshel in the 1970s, we now have powerful tools that combine classical and quantum approaches.

We can study complex molecular systems involving molecules of unlimited size by focusing the appropriate (most demanding) level of study on the relevant part of the system, applying a lighter level to the rest. For example, in the models built to study drug-target interactions, complexes are analyzed in which the drug is introduced into the binding site of the target, designing a strategy in which quantum calculations are applied to the drug and to the section of the target contained within a given radius around it (the most interesting approach for analyzing the intra- and inter-molecular interactions of the system), while classical calculations are applied to the rest, the part of the target farthest away from a given point. Thus, with reasonable time and calculation consumption, more real information on the binding mode is obtained, which facilitates the strategies of design based on the structure of the therapeutic target.

The versatility of these programs opens the door to their application to countless areas of interest.