Lect2-Modeling

=Lecture 2- Models, Modeling and Definitions=

Chemistry deals with the properties of matter and with their changes. We attribute these properties to the motions of electrons and atomic nuclei, but this is just one particular model for the universe we occupy. There are a number of theories we can use to explain the behavior of matter and these include:

Quantum mechanics: the mechanics of atoms and their combination in molecules. Classical mechanics is just a special case of quantum mechanics.

Statistical mechanics: the framework by which molecular properties can be related to the macroscopic behavior of chemical substances.

Thermodynamics: the study of energy and order-disorder, and their connections with chemical changes and chemical equilibrium.

Kinetics: the study of the rates of chemical reactions and of the molecular processes by which reactions occur.

These are the fundamental subjects of physical chemistry and of molecular modeling. Most any process we want to model requires some calculation/estimation of the forces between atoms and/or molecules (think docking, folding, solvation) and this can be done with either quantum or classical mechanics, the rate and probability of the process occurring (binding of a ligand is a competition between the rate of binding and the rate of release) and the equilibrium constant measured is just the macroscopic variable.

On a more practical level, chemists use physical models at every level of study. We make students purchase a set of plastic models in first year chemistry and have them build and manipulate the models of different structures. The intent is to impart in the students an idea of the three dimensional nature of molecules, the relative size and strengths of various bonds. Of course these models are static – the bonds don’t vibrate, and the atoms don’t rotate about the bonds and so the picture is very limited and biased. It is very difficult to go from a static model to a reaction and most students never make the connection. These physical models have their uses – questions about distances between atoms can easily be discussed (myoglobin model) but they are not very realistic.

A similar situation occurs in computer modeling – the user builds/inputs a structure of interest and then optimizes the geometry and gets back a static structure. She can then query the structure for various properties (dipole moment, solubility, logP, nmr coupling constants), manipulate the structure much as she did with the physical model. This computer model may be more accurate in terms of bond lengths and angles but it still does not convey the dynamic nature of the molecule. I go over this because while modeling is good and useful you need to keep reminding yourselves of the limitations of your models. Models are not real – they are tools we use to help us think about past experiments and to guide us in doing new experiments.

We have talked about experiments and theories, ideas that we use to help rationalize the experiments. Where does computer modeling fit – it can be thought of as theory reduced to practice but computer simulations can also be viewed as experiments in their own right. For this class:

Model is a simplified or idealized description of a system or process devised to facilitate calculation and prediction.

Molecular modeling is any theoretical or computational technique that provides insight into molecular systems.

Course of Study: Compute forces between atoms and molecules: Quantum Mechanics Classical Mechanics Minimize the Forces Optimization methods Search Conformational Space Small molecule search methods Molecular Dynamics Monte Carlo Mesoscale modeling