Proteins are polypeptide chains that do fold to adopt one compact structure, while polymers do not. Understanding this mechanism is a great deal for the biologist. It involves all fields of protein engineering, i.e. the ab initio construction of a protein bearing a particular biological function one is interested in. The aim is to be able to cover as fast as possible the distance between sequence and structure, with the highest possible accuracy if one wants to achieve biological function.
The total number of sequences, of the order of 10^10 or 10^12, collapses at the three dimensional level in a fairly limited number of folds, i.e. of different topologies, probably of the order of one thousand. At the time of automatic sequencing of entire genomes, sequence information is available at the molecular level, but typically one quarter of the genes of new genomes consist of unknown functions and much more have unknown structure. To be able to predict the folding family by the knowledge of the sequence alone is an important step because it allows to provide a function to a new gene.
If one wishes to predict the folding, regardless of the intermediate states and of the mechanism, many techniques based on mathematics or statistics are used (e.g. neural networks or genetic algorithms) in the purpose of predicting secondary structures.
On the other hand, the one of the mechanism itself, emerging spectroscopies give interesting insights, as for instance 2D Fourier transform infrared spectroscopy or pressure jump.
It will be the time, in 2001, to review these new experimental methods and check if their promises have accomplished.