Common protein surface shapes and uses therefor

Abstract

A method of determining common three-dimensional structural features of protein surfaces is provided, as is use of representations of these common structures in molecular database searching and in designing focussed molecular libraries. The method is particularly concerned with the analysis and representation of protein surfaces such as b-turns, loops and contact surfaces. In one form, the method identifies common locations and orientations of amino acid side-chains, simplified as Cα-Cβ vectors. In another form, the method identifies common regions of surface charge represented by grid points in three-dimensional space. Further provided are common three dimensional structural features of proteins that can be used to search molecular databases for the purposes of identifying molecules that match these common three dimensional structural features. The common three dimensional structural features can also be used to focus de novo molecular generation to produce libraries containing molecules that have these common three dimensional structural features. Libraries of these structurally-related molecules may then be produced for the purposes of drug discovery.

1 . A computer-implemented method of producing a description of a common three-dimensional protein surface shape including the steps of: 
 (i) identifying a three-dimensional surface shape of each of a plurality of proteins; and    (ii) creating one or more descriptors wherein each said descriptor represents a common surface shape derived from respective said three-dimensional surface shapes of two or more proteins of said plurality of proteins.    
   
   
       2 . The method of  claim 1 , wherein each three-dimensional surface shape identified at step (i) is represented by a side-chain location and orientation of two or more amino acids of each surface.  
   
   
       3 . The method of  claim 2 , wherein at step (ii) each said descriptor represents a common location and orientation of respective side chains of two or more amino acids of each of said two or more proteins.  
   
   
       4 . The method of  claim 3 , wherein the or each said descriptor represents a common location and orientation of respective side chains of three or more amino acids of each of said two or more proteins.  
   
   
       5 . The method of  claim 3 , wherein the or each said descriptor represents the common location and orientation of respective side chains of two or more amino acids of each of three or more proteins.  
   
   
       6 . The method of  claim 2  or  claim 3 , wherein the location and orientation of each said amino acid side chain is in three-dimensional (3D) space.  
   
   
       7 . The method of  claim 6 , wherein each amino acid side chain used to produce said descriptor is simplified as a Cα-Cβ vector.  
   
   
       8 . The method of  claim 1 , wherein each three-dimensional surface shape identified at step (i) is represented as a charged surface region of each said protein.  
   
   
       9 . The method of  claim 8 , wherein at step (ii) each said descriptor represents a common charged surface region of two or more proteins of said plurality of proteins.  
   
   
       10 . The method of  claim 9 , wherein each charged surface region is represented by at least four grid points.  
   
   
       11 . The method of  claim 10 , wherein respective said grid points are 0.2 to 2.0 angstrom apart in three dimensional (3D) space.  
   
   
       12 . The method of  claim 11 , wherein respective said grid points are 0.5-1.5 angstrom apart in three dimensional (3D) space.  
   
   
       13 . The method of  claim 9 , wherein the or each said descriptor represents the common charged protein surface shape of three or more proteins.  
   
   
       14 . The method of  claim 1 , wherein said three-dimensional surface shape is of at least part of a structural feature of each of said two or more proteins.  
   
   
       15 . The method of  claim 14 , wherein said structural feature is, or comprises, a β-turn, a loop or a contact surface.  
   
   
       16 . The method of  claim 14 , wherein said structural feature is, or comprises, a loop or a contact surface.  
   
   
       17 . The method of  claim 16 , wherein the contact surface comprises one or more discontinuous and/or continuous surfaces.  
   
   
       18 . The method of  claim 3 , wherein said descriptor represents side-chain location and orientation four β-turn or loop amino acids.  
   
   
       19 . The method of  claim 3 , wherein said descriptor represents side-chain location and orientation of at least three amino acids of a contact surface.  
   
   
       20 . The method of  claim 19 , wherein said descriptor represents side-chain location and orientation of four, five, six or seven amino acid side-chains of a contact surface.  
   
   
       21 . A computer-implemented method of identifying one or more molecules having a common three-dimensional protein surface shape, said method including the steps of: 
 (i) creating a query using one or more descriptors that each represent a common three-dimensional protein surface shape; and    (ii) using said query to search a database and thereby identify one or more entries in said database that correspond to one or more molecules that each match said descriptor.    
   
   
       22 . The method of  claim 21 , wherein at step (i), the descriptor represents a common amino acid side-chain location and orientation of two or more amino acids of each of two or more proteins.  
   
   
       23 . The method of  claim 21 , wherein at step (i), the descriptor represents a common protein surface charge shape of two or more proteins.  
   
   
       24 . The method of  claim 21 , wherein at step (i), the query comprises: 
 (a) a descriptor that represents a common side chain location and orientation of two or more amino acids of each of two or more proteins; and    (b) a descriptor that represents a common protein surface charge shape of said two or more proteins.    
   
   
       25 . The method of  claim 21 , wherein each amino acid side chain used to produce said descriptor is simplified as a Cα-Cβ vector.  
   
   
       26 . The method of  claim 25 , wherein the Cα-Cβ vectors of said query are represented as a distance matrix.  
   
   
       27 . A computer-implemented method of creating a library of molecules including the steps of: 
 (i) searching a database to identify one or more entries corresponding to one or more molecules that each match a common protein surface shape; and.    (ii) using at least one of the one or more molecules identified at step (i) to create a library of molecules.    
   
   
       28 . The method of  claim 27 , wherein said library of molecules is a virtual library.  
   
   
       29 . The method of  claim 27  said library of molecules is a synthetic chemical library.  
   
   
       30 . A method of engineering one or more molecules including the steps of: 
 (i) creating one or more descriptors that each represent a common three-dimensional protein surface shape; and    (ii) engineering one or more molecules that respectively comprise one or more structural features according to the or each descriptor in (i).    
   
   
       31 . The method of  claim 30 , wherein the or each descriptor represents a common side chain location and orientation of two or more amino acids of each of two or more proteins.  
   
   
       32 . The method of  claim 30 , wherein the or each descriptor represents a common protein surface charge shape of said two or more proteins.