Molecular modeling

In my opinion molecular modeling is a discipline at the borderland of science, technology and arts.

1tn2 Molecular modeling consists in creating models of chemical molecules, which help to discribe or predict their real physical/chemical properties. As a "patron saint" of molecular modeling we could recognize André Kekulé who, on the grounds of benzene's properties and composition, determined (correctly!) its structure. According to a legend he did it imagining a model of six monkeys holding their hands and legs and forming a ring.

Later on, models were built from metal or plastic to assist our imagination in noticing features lost in two-dimensional sketches. A CPK (or close-packed sphere) model of the yeast phenylalanine tRNA molecule is shown on the left. It is represented on a flat screen but it resembles such plastic models.

Unfortunately, plastic can simulate rather few features of a real molecule; in fact only its enlarged three-dimensional structure.

Advancements in computer technology and development of molecular physics made it possible to create models which describe interactions between atoms and molecules. Therefore simulations of the behavior of real molecules on the grounds of physical laws are now feasible. Scientists can now create mathematical systems (so-called force field algorithms) which reproduce the interactions between various types of atoms found in nature. This is called molecular mechanics. Properties of small molecules can also be predicted using the quantum mechanics formulas. This approach provides more exact data but requires vast computational capacities.

Using professional programs and supercomputers one can, for instance, search for the most stable structure of a given molecule (so-called energy minimization) or simulate its oscillations at any given temperature (molecular dynamics). These techniques are useful for the refinement of conformations of, for example, protein molecules determined by physical methods (as X-ray crystallography or nucleic magnetic resonance). It is far more challenging to employ these methods to predict the conformation, dynamics and properties of molecules whose structure is not yet determined by independent methods.

At the same time, professional graphic workstations, as Silicon Graphics IRIS Indigo 2 (the one I have been mostly using in my research) make perfect visualization of molecules possible. On the adjacent picture, nucleobases - components of nucleic acids - are presented on the screen of an IRIS Indigo 2 workstation.

3cro The techniques listed above make it possible to predict the structure and properties of molecules, which haven't been even synthetized yet - this is called molecular design. Molecular design can be applied in the search for new materials, drugs etc., where one has to consider all possible modifications of known chemical substances, a task that might be too expensive or difficult for traditional methods. Molecular modeling provides a possibility of screening for those modifications which are most likely to be successful.

 Molecular modeling is very useful in predicting the alignment of molecules in complexes. The image on the left represents a DNA (violet) complex with the cro protein. It's important in designing new drugs which have to interact in a desired way with, for instance, cell receptors.

In my research I have long been using the Accelrys InsightII program which is part of the MSI software. My calculations are performed using the AMBER software running on Cray and Silicon Graphics supercomputers which are maintained by the Poznan Supercomputing and Networking Centre.

The computer graphics available today help in visualization of protein and nucleic acid molecules in various ways, so as to emphasize their important structural and functional features.

9pti  A model of bovine pancreatic trypsine inhibitor (BPTI) is presented on the left. The ribbons show the course of the main chain of the protein whereas the thin lines depict the side groups of the aminoacids. Different colors mark fragments with different secondary structure, e. g. the helical segments are colored red.

 Shown on the right is a DNA complex with the diaminophenylindol dye (DAPI) represented as a color coded sphere model. The DAPI molecule is grey whereas the different types of DNA bases are marked with different colors. The dye molecule is bound in the minor (narrow) groove of the DNA duplex, which explains its ability to associate with DNA and, consequently, to stain it.


Details about my own research are also available.

mailbox Comments, remarks or questions

BackReturn to the Lukasz Bielecki's homepage space PLWersja polska .