8.3 Protein folding and secondary structure prediction

Proteins are the biological molecules that are the building blocks of cells and organs, and the biochemical processes required to keep living organisms alive are catalyzed and regulated by a particular category of proteins called enzymes. Proteins are linear polymers of amino acids that fold into complex conformations dictated by the physical and chemical properties of the amino acid chain. The biological function of a protein is dependent on the protein folding into the correct, or "native", state. Protein structure is described by biologists in terms of primary structure, which is the amino acid sequence, secondary structure, wherein the polypeptide backbone assembles into local regions of alpha-helices, beta-sheets, coils and turns, tertiary structure, which refers to the entire 3-dimensional structure of the protein, and quaternary structure, which describes interactions between separate polypeptide chains, called subunits, that exist in some large protein complexes. Computational methods have been developed that can predict protein secondary structure with a reasonable degree of accuracy. Prediction methods exist for predicting tertiary structure, but the accuracy of such methods is highly dependent on whether or not the protein in question is related in sequence to any members of the existing library of known protein structures. The development of ab initio tools to predict the complete structural fold of a protein from its amino acid sequence is a burgeoning field in computational biology, but true attainment of this goal is still pretty distant.

Protein folding is usually a spontaneous process, and often when a protein unfolds because of heat or chemical denaturation, it will be capable of refolding into the correct conformation, as soon as it is removed from the environment of the denaturant, meaning folding and unfolding under these circumstances are reversible. Protein folding can go wrong for many reasons. When an egg is boiled, the proteins in the white unfold and misfold into a solid mass of protein that will not refold or redissolve. In a similar way, irreversibly misfolded proteins form insoluble protein aggregates found in certain tissues that are characteristic of some diseases, such as Alzheimer's Disease.

Determining the process by which proteins fold into particular shapes, characteristic of their amino acid sequence, is commonly called "the protein folding problem". One approach to studying the protein folding process is the application of statistical mechanics techniques and simulations to the study of protein folding. (1) These methods allow the investigation of larger systems than methods that try to represent atomic detail in their simulations of biological molecules, and have had success correlating the computational folding model with folding intermediates and transition states that have been experimentally measured for a limited test set of relatively large proteins.