Molecular modelling

Molecular modelling is a blanket term used to describe the branch of chemistry in which scientists simulate and experiment with the physical structures and chemical properties of molecules. Researchers use these simulations to predict reactivity and changes in chemical structures brought about by interactions among various molecules and materials. Today, molecular modelling usually requires the use of specialized computer software.

Computer-based molecular models are often developed in three steps. In the first step, scientists define the preliminary characteristics of the molecules being modelled, including the number of atoms they possess and the ways in which coexisting elements are joined or bonded. Next, the molecules are represented in three dimensions by establishing the relative positions of the constituent atoms. Finally, chemical and molecular interactions are simulated, and scientists log and compare the characteristics of the resultant materials against their initial predictions.

Applications

Molecular modelling is widely used in pharmaceutical research, particularly in drug design and refinement. It also has important theoretical applications in computational chemistry, computational biology, and materials science, a field that seeks to discover and develop new synthetic substances. Scientists involved in bioinformatics, which is concerned with the analysis of biochemical data, also use molecular modelling techniques.

Specific Molecular Modelling Methods

In the natural world, molecular connections are largely governed by a process known as molecular recognition. This process occurs in both biological and inorganic systems and describes the means by which various molecules interact, both directly and indirectly. Some interactions result in the creation of new complexes from two or more molecules. Predicting how these molecules will bind together is known as docking. At their most basic level, specific molecular modelling techniques are used to determine if and how docking takes place. If it does, these techniques can also estimate how strong the resultant chemical bonds or binding affinity will be.

At room temperature, the atoms in most molecules have sufficient kinetic energy to continue moving in ways that are not always easily predictable. Thus, the first objective of any molecular modelling technique is to create an accurate representation of how the molecules being studied would behave under such conditions, prior to any molecular interactions. This enables researchers to develop precise and reliable interaction models.

The dynamic nature of docking also means that different molecular bonds can form as the result of an interaction; some bonds are weak and momentary, while others are strong and long lasting. Therefore, another primary objective of molecular modelling is to identify and test the strength of the ensuing binding affinities.

Specific modelling techniques are chosen and applied based on the size and characteristics of the system being simulated. Asymmetrical and very large systems are usually modelled using an approach known as molecular mechanics, while a system called molecular dynamics is favored in cases where the molecules have a high level of innate kinesis. A range of highly specialized techniques has also been developed for the modelling of systems with rare, complex, or extremely specific attributes.

In most cases, unusually large and nonsymmetrical chemical systems, such as those seen in many polymers and proteins, can only be properly and accurately modelled using molecular mechanics. This approach is heavily reliant on empirical observation, and it applies the laws and principles of classical physics to forecast the chemical makeup of molecules created through interactions.

Molecular mechanics has several important limitations. It cannot predict or quantify bonding affinities, as these take place on a quantum level beyond the scope of the technique. In addition, molecular mechanics energy calculations are entirely dependent on the system in which they originated. In other words, these calculations cannot be objectively understood outside of that system. As such, they are only useful in applications involving comparative research. Despite these limitations, molecular mechanics benefits from the high degree of specificity used to set experimental parameters. This has led to superior accuracy in the generation of geometric values for the vast majority of experimental organic molecules.

In molecular dynamics, organic molecules, which are prone to movement, are infused with an additional burst of kinetic energy then left to interact for a fixed period while scientists observe their evolution. This approach can be effectively combined with supplementary techniques, such as minimization phasing, to allow for a more detailed and comprehensive study of molecular behavior at various temperature ranges.

Beyond these two techniques, researchers have developed a number of specialized approaches for measuring binding affinities and studying highly specific molecular interactions, especially those that involve organic compounds. The following are examples of some of these approaches.

• AMBER (Assisted Model Building with Energy Refinement): The inherent parameters of the AMBER system make it ideal for modelling nucleic acids and organic proteins.
• CHARMM (Chemistry at HARvard Macromolecular Mechanics): Built around an electrostatic force field system, the CHARMM technique was originally developed to test proteins and nucleic acids. It has since been expanded and refined for use in a broader range of applications, including quantum mechanics/molecular mechanics, crystal packing, molecular dynamics, and vibrational analysis.
• CVFF (Consistent Valence Force Field): Used in the modelling and analysis of gas phase structures and organic crystals, the CVFF system was specifically created to study and predict the binding affinities of indicated molecules. It can also accurately predict post-interaction characteristics, including conformational energy and vibrational frequency.
• GROMOS (Groningen Molecular Simulation): This technique, developed in Switzerland and the Netherlands, has proven very useful in the molecular-level study of bulk liquids. It is also used for the modeling of molecular structures found in living organisms.
• MMFF (Merck Molecular Force Field): The Merck Molecular Force Field technique is versatile, with a relatively wide range of useful applications, especially in the fields of organic chemistry, medicinal chemistry, and geometrical optimization.
• OPLS (Optimized Potential for Liquid Simulations): Like GROMOS, OPLS is usually used for applications involving bulk liquids and biomolecules.

Researchers are constantly refining existing techniques and creating new ones as the limits of the computing technologies used in molecular modelling continue to expand. As such, the field itself is highly dynamic and benefits from the regular emergence of new applications.

Bibliography

Bladon, Peter, John Gorton, and Robert B. Hammond. Molecular Modelling: Computational Chemistry Demystified. Cambridge, UK: Royal Society of Chemistry, 2011, 1–10. Print.

Hinchliffe, Alan. Molecular Modelling for Beginners. Chichester, UK: John Wiley & Sons, 2011. Print.

Morris, G.M., and M. Lim-Wilby. Molecular Docking. PubMed. National Center for Biotechnology Information, U.S. National Library of Medicine. Web. 30 Dec. 2014. http://www.ncbi.nlm.nih.gov/pubmed/18446297

Yunta, María J. R. "Using Molecular Modelling to Study Interactions between Molecules with Biological Activity." Intech. Intech. Web. 29 Dec 2014. http://www.intechopen.com/books/bioinformatics/using-molecular-modelling-to-study-interactions-between-molecules-with-biological-activity