Method for determining three-dimensional structures of dynamic molecules

Abstract

The present invention relates to a method for determining three-dimensional structures of molecules, particularly, but not exclusively, dynamic organic molecules of biological interest such as peptides, carbohydrates, proteins and drug molecules. A first aspect of the present invention provides a method for generating data representing an ensemble of three-dimensional structures of a molecule, the molecule comprising first and second atoms linked by at least one bond, said bond having an associated angle, and the angle varying to generate a plurality of three-dimensional structures of said molecule, the method comprising: receiving data representing said molecule, said data comprising data indicating variability of said angle; and generating an ensemble of structures such that the angle has an associated value selected based upon said variability. A second aspect of the present invention provides a computer implemented method for simulating the variability of the three-dimensional structure of a molecule.

1 - 62 . (canceled) 
     
     
         63 . A method for generating an ensemble of discrete molecular structures representing a range of three-dimensional shapes of a solvated molecule, the molecule comprising a plurality of bonds each having an associated mean dihedral angle, wherein at least one bond is a rotatable bond having an associated distribution of dihedral angles, the method comprising:
 generating an instance of a flexible molecule data struct;   generating a plurality of instances of a bond data struct, wherein each of the plurality of instances of the bond data struct are linked with the instance of the flexible molecule data struct and wherein each of the plurality of instances of the bond data struct include a predefined mean dihedral angle for a bond;   assigning a probability distribution function to each of the plurality of instances of the bond data struct, each probability distribution function comprising a predefined variability of the dihedral angle for a bond with respect to said predefined mean dihedral angle for the bond, wherein the probability distribution function is independent of experimental data for the solvated molecule;   iteratively generating, with a computer, an ensemble of data structs that represent the ensemble of discrete molecular structures by iteratively implementing the flexible molecule data struct and varying said dihedral angle for each bond data struct according to said predefined dihedral angle variability for each iteration; and   analyzing the ensemble of discrete molecular structures as a group,   wherein the predefined variability of the dihedral angle for a rotatable bond is non-zero and corresponds to the associated distribution of dihedral angles.