GEPOL cavities

Figure 3. GePol cavity for -alanine zwitterion subdivided into tesserae with average area of 0.4 and of 0.2 respectively.

The cavity surface is then subdivided into small domains, called surface tesserae, used to express as finite sums all the surface integrals needed to compute the solvent reaction field, as explained below. The final result is depicted in figure 3, where we show the GePol cavity for p-alanine subdivided into tesserae with average area of 0.4 and of 0.2 respectively. 

Define a grid of points at the cavity surface. This can be done using the Gepol program for van der Waals-type surfaces (see below) [60,61],... 

Unlike many other continuum approaches, PCM adopts cavities of realistic shape modelled on the solute atoms they are built according to GePol algorithm, [114] in which the cavity is defined as the envelope of spheres centred on solute atoms or atomic groups. Besides the atomic spheres, other spheres are added by GePol to smooth the solute-solvent boundary, approximating the so-called solvent accessible surface proposed by Connolly. 

The first point becomes important when thousands of atomic spheres have to be defined and cross-checked to eliminate the parts of the surface lying inside the cavity in the case of globular proteins, for exanq)le, more than 90% of the generated spheres are eventually discarded since they are not in contact with the solvent but they still contribute to the time spent to buUd the cavity. If all or a large part of the solute is treated at the MM level, the GePol procedure is the... 

Figure 4. GePol and DefPol cavities for the alanine tripeptide in a-helix conformation.

Formulas are given in the already quoted Cossi et al. s paper (1996a). To compute this contribution with Pierotti-Claverie s formula (eq. 80), the tessellation of the cavity portions is not compulsory. However, to exploit the subroutines used for G%is and G j, we can use the partial surface calculations via summation over the tesserae. Also the total volume is computed by using a BEM procedure, i.e. exploiting the definition of tesserae. According to the definition of the SPT cavity, there is no need of introducing GEPOL additional spheres (the Sas surface is here used). The computational cost of G av is minimal. 

We recall that in standard PCM we exploit the GEPOL definition of the cavity and of its tessellation [46, 47, 48] with a more accurate definition of the Mea of those tesserae which are cut by other spheres (the so-called Gauss-Bonnet partition, GEPOL-GB [31]). Other definitions of the molecular cavity and its tessellation will be given later. 

We have recalled that in PCM molecular cavities there can be additional spheres inserted to describe solvent excluded space. The radii of such spheres can have very different values, but they are generally small. In the GEPOL... 

---