Polypeptide force field

The preceding discussions have established the theoretical basis for computing the normal modes of vibration of a polypeptide chain molecule. Throughout the discussion we have assumed knowledge of the structure of the molecule and of its vibrational potential energy function. It is now necessary to examine these two kinds of inputs, and in particular to understand how we can obtain a polypeptide force field that might serve to predict the vibrational frequencies of an arbitrary chain conformation. 

Figure 4-1. Schematic illustration of the partition between two residues along a polypeptide chain in the X-pol force field...

The computational verification and prediction of the conformations (tertiary structures) of polypeptides presents a considerable challenge for theoretical chemists. Attempts towards this goal have so far made use of force field calculations to a large extent. Levitt... 

There is considerable activity in the area of MM calculations on small polypeptides. A preliminary amide force field based on the CFF (231,232) and a partial relaxation force field (233) has been reported. Interested readers are referred to review articles cited in ref. 232. 

Hall D, Pavitt N (1984) An appraisal of molecular force fields for the representation of polypeptides, J Comp Chem, 5 441—450... 

The large body of literature on calculations of structures of proteins, polypeptides, polysacharides and polynucleotides is not reviewed. These calculations, in so far as they apply to the force-field approach, are necessarily based on highly simplified fields (hard- or soft sphere approach) and furthermore, due to limits in computer memory capacity and speed, full relaxation of the atomic coordinates of such large molecules (> 75 atoms) is as yet unattainable. 

The development of a force field suitable for the polypeptide chain involves several steps. First, as we will see below, it is necessary to select a physically appropriate form for the potential, V, both for the intramolecular, Vintra, and the intermolecular, Vimer, parts. Second, we need a formalism for producing an acceptable set of force constants from the observed data this is usually done by a least-squares procedure. Third, the molecules used, and their sequence, in the refinement process... 

In analyzing the vibrational spectra of polypeptides, it is important to include certain intermolecular contributions to Vinter- We discuss here three of these contributions, two of which, hydrogen bonding and transition dipole coupling, have played a very important role in the development of our force field. 

In developing a transferable SGVFF for the polypeptide chain, we have utilized strategies based on all of the considerations discussed thus far in this article. We review these developments before presenting the details of the force field. 

In a subsequent analysis, the force field was further refined for maximum transferability between different polypeptide molecules (Dwivedi and Krimm, 1982a). The following discussion is based on these two papers (Moore and Krimm, 1976a Dwivedi and Krimm, 1982a). 

---