First-principles Methods for Predicting Protein Structure

3 First-principles Methods for Predicting Protein Structure 

The most ambitious approaches to the protein folding problem attempt to solve it from first principles ab initio). As such, the problem is to explore the conformational space of the molecule in order to identify the most appropriate structrue. The total number of possible conformations is invariably very large and so it is usual to try to find only the very lowest energy structure(s). Some form of empirical force field is usually used, often augmented with a solvation term (see Section 11.12). The global minimum in the energy function is assumed to correspond to the naturally occurring structure of the molecule. 

All of the conformational search methods that were described in Sections 9.2-9.7 have been used at some stage to explore the conformational space of small peptides. Here we will describe some of the methods designed specifically for tackling the problem for peptides and proteins, 

One reason for the interest in lattice models is that they can be used to try to answer some of the fundamental questions about protein structure. For example, it may be feasible to enumerate all possible conformations for a chain of a given length on the lattice From this set of states statistical mechanics can be used to derive thermodynamic properties and to investigate the relationship between the structure and the sequence. In the HP model [Chan and Dill 1993], a protein is modelled as a sequence of hydrophobic (H) and 

To summarise, first-principles methods have been successfully used to predict the naturally occurring conformations of small peptides but are not yet sufficiently reliable to predict accurately the structures of proteins, though in some cases the general fold of the molecule is quite similar to the native structure. However, as we shall see later in Section 10.8 lattice models have been very useful in helping to understand protein folding. 

---