Home » Technology » Scientific Computing » Simulating Biological Building Blocks

Simulating Biological Building Blocks

© Adam Hughes

In order for researchers to be able to conduct meaningful computer simulation studies of systems with important biological implications, such as DNA, large protein structures, and steroids, a firm understanding of basic interactions in hydrocarbons and other small pre-biological molecules is needed. Many decades of research have been spent in these endeavors, and the results leave us on the brink of a new age of understanding.

As with other chemical simulations, the starting point in studying biological systems lies in quantum chemistry. Recall that quantum chemistry involves treating a molecule's electrons in some way. The most rigorous form of this kind of calculation is ab initio, literally, "from the beginning". This involves treating all electrons explicitly, generating inter-electron energies and forces, and using the resultant force field to move the atoms of the system to some minimum energy. This type of calculation is very expensive but also very accurate. It is suitable for small molecules, and has been used to gain an understanding of bond stretching, bending, and rotating energies between the carbon, nitrogen, hydrogen, and oxygen atoms that typically comprise a biological molecule.

Beyond ab initio studies, there are semi-empirical methods, which treat some electrons in the system explicitly, but take other interactions to be parameters which are set once and remain unaltered for the remainder of the simulation. These parameters may come from experimental studies or more accurate quantum chemistry studies. Because there are less electronic interactions to compute, a semi-empirical calculation will be less computationally costly than a corresponding ab initio calculation on the same system. As a result, it is possible to examine larger systems with the computational resources available. These methods have allowed scientists to study larger molecules, such as amino acids and carbohydrates with quantum methods. Typically, various energies and structural properties can be calculated in this manner.

While the quantum methods can give very accurate results when the goal is learning spectroscopic data or structural properties of a molecule, many of the important properties emerging today involve large motions of molecules or systems of molecules. To study these types of situations, scientists can turn to molecular dynamics or monte carlo situations. These methods employ empirical or semi-empirical force fields, derived either from experiments or quantum calculations. These methods lend themselves to studying large systems, or smaller systems for long periods of time. Typical applications include the study of proteins, liquid hydrocarbons, and intricate crystal structures.

In addition to these methods, there exist hybrid technologies which allow the researcher to generate his own, on-the-fly potential energy surfaces (quantum calculation), and then move the atoms accordingly (classical dynamics). There are also innovations being developed all the time which allow bigger

The copyright of the article Simulating Biological Building Blocks in Scientific Computing is owned by Adam Hughes. Permission to republish Simulating Biological Building Blocks in print or online must be granted by the author in writing.