Energy-based methods

New Free Energy Based Methods for Ligand Binding from Detailed Structure-Function to Multiple-Ligand Screening... 

S. Banba, Z. Guo, and C. L. Brooks III, New free energy based methods for ligand binding from detailed structure-function to multiple-ligand screening, in Free Energy... 

Laurie ATR, Jackson RM (2005) Structural bioinformatics Q-SiteFinder an energy-based method for the prediction of protein—ligand binding sites. Bioinformatics 21 1908-1916... 

GRID and MCSS use classical mechanics potential functions for scoring and selecting the fragments, and are usually called energy based methods [37]. 

Another energy-based method looks at the potential of mean force for the p-carbon atoms. Using a database of known protein structures, the potential of mean force for all amino acid pairs was compiled. The conformational energy of sequences was then calculated for a number of different folds, and it was found that in most cases the native state had the lowest energy. Most of the exceptions were structures with large prosthetic groups, Fe—S clusters, or nonglobular proteins. 

An alternative to the energy-based methods such as GRID is to suggest possible binding positions using a knowledge-based approach. Analysis of experimentally determined... 

The last step in the refinement process is to apply an energy based method as molecular mechanics and dynamics simulation to obtain a model that is energetically stable. In the absence of a membrane fhaf keep fhe helices together, it is advisable to introduce a small force consfanf fo keep the backbone of fhe transmembrane segments in place and to avoid gross artificial disforfions. This force consfanf can be reduced in subsequenf cycles of dynamics simulation (equilibration followed by dynamics) in gradual steps unfil completely eliminated. 

This uses the geometric methods for searching the active sites different from the energy-based methods used by Schrodinger. In this, the relative positions and accessibility of receptor atoms are considered. We use the same protein example as above... 

Knowledge-based methods can clearly provide valuable predictive information and will become increasingly powerful as the quantity of relevant data (i.e., the number of known crystal structures) increases. These methods are not an alternative to the continued development of lattice-energy-based methods, but are complementary methods that provide further information that can be used to assess the likely crystal forms of a molecule. 

This energy-based method of calculation of viscosity rj is due to Einstein [87], who considered hydrod)mamic dissipation in a very dilute suspension of non-interacting spheres. Tanaka and White [86] base their calculations on the Frankel and Acrivos [88] cell model of a concentrated suspension, but use a non-Newtonian (power law) matrix. The interaction energy is considered to consist of both van der Waals-London attractive forces and Coulombic interaction, i.e. 

The majority of the sequential buildup procedures use energy-based methods for fragment scoring, although several of the methods use a rule-based approach. 

This approach is amenable to carrying out Monte Carlo simulations that employ the relative free energy A< > to build the Markov chain. With some care when extracting the bulk properties of the system, this method also provides reliable results for a range of conditions. However, these results deteriorate with increasing diffraction effects, as shown by comparison with exact PIMC results [103]. The reader is referred to Ref. 145 for specific details about this free-energy-based method. 

---