|
|
||
MOLECULAR MODELINGIn a pool table after a cue exerts a force on a ball many balls begin to move as a result of interaction between the balls. If we know the position of each ball using Newton's laws we can predict the motion of all balls with reasonable accuracy, if we simplify the problem to a certain extent (3 body problems have no solutions otherwise). In order to solve such a problem computationally we have to make many simplifications such as assume elastic collisions, and so on.
Many of the problems that we would like to tackle in molecular modeling are unfortunately too large to be considered by quantum mechanical methods even for distributed computing. As protein models consist of hundreds or thousands of atoms the only feasible methods of computing systems of such size are molecular mechanics calculations (kind of like playing pool with atoms.) A force field is assigned to each atom in the protein. This figure is a schematic representation of the four key contributions to a molecular mechanics force field: bond stretching, angle bending, torsional terms and non-bonded interactions.
Effects of water
on proteins is also integrated into force fields by various methods.
Once force fields are assigned to each atom, then computers do computations
to follow the motion of the atoms. Until recently, it was thought these
computations were not feasible in the study of protein folding. Pande
group with the novel distributed
computing approach successfully simulated folding of many proteins.
Once the simulation is done, much valuable information can be derived
from the data. With data
like these, scientists hope to find the cure for many diseases like
Alzheimer's thought to be caused by improper folding of proteins. Author: Tug Sezen Reference: Andrew R. Leach (1996); Molecular Modeling - Principles and Applications; Addison Wesley Longman Limited
|
