Prediction of molecular polar surface area and bioabsorption

Abstract

The polar surface area of molecules is computed without reference to three dimensional molecular structures using a linear equation incorporating counts of nitrogens, oxygens, and related atom clusters. Methods and systems for predicting intestinal absorption of candidate compounds use the polar surface area and the octanol/water partition coefficient as descriptors.

What is claimed is:  
     
         1 . A method of estimating the polar surface area of a molecule comprising: 
 making separate counts of the number of atoms or atom clusters in said molecule which fall within the definition of each of a plurality of atomic classes;    estimating a polar surface area of said molecule using at least some of said separate counts.    
     
     
         2 . The method of  claim 1 , wherein said estimating comprises multiplying at least some of said counts by one or more coefficients, and summing the products thereof.  
     
     
         3 . The method of  claim 1 , wherein said plurality of atomic classes includes one or more of single bonded NH 2 , double bonded NH, doubly single bonded NH, aromatically bonded NH, aromatically bonded N, triple bonded N, single bonded OH, double bonded O, doubly single bonded O, and aromatically bonded O.  
     
     
         4 . The method of  claim 3 , wherein said plurality of atomic classes includes all of single bonded NH 2 , double bonded NH, doubly single bonded NH, aromatically bonded NH, aromatically bonded N, triple bonded N, single bonded OH, double bonded O, doubly single bonded O, and aromatically bonded O.  
     
     
         5 . A method of estimating the molecular polar surface area of a molecule comprising performing counts of pre-defined atom types and/or atom cluster types present in said molecule and estimating said molecular polar surface area using said counts weighted by coefficients and summed together, without reference to an energy minimized three-dimensional molecular structure of said molecule.  
     
     
         6 . The method of  claim 5 , wherein said pre-defined atom types and/or atom cluster types includes one or more of single bonded NH 2 , double bonded NH, doubly single bonded NH, aromatically bonded NH, aromatically bonded N, triple bonded N, single bonded OH, double bonded O, doubly single bonded O, and aromatically bonded O.  
     
     
         7 . A method of predicting the propensity of a molecule for membrane permeability comprising: 
 computing a polar surface area for said molecule without reference to an energy minimized three dimensional structure of said molecule; and    using said computed polar surface area to predict said propensity.    
     
     
         8 . A method of drug discovery comprising: 
 computing polar surface areas for a plurality of molecules without reference to an energy minimized three dimensional structure for at least some of said molecules; and    using said computed polar surface area in a membrane permeability prediction model so as to select one or more of said molecules for further analysis.    
     
     
         9 . The method of  claim 8 , additionally comprising computing logP for said plurality of molecules, and using such computed logP in said membrane permeability prediction model.  
     
     
         10 . The method of  claim 8 , comprising prioritizing compounds for synthesis and screening.  
     
     
         11 . A computer readable medium having instructions stored thereon which cause a general purpose computer to perform a method of molecular PSA estimation, said method comprising: 
 making separate counts of the number of atoms or atom clusters in a molecule which fall within the definition of each of each of a plurality of atomic classes;    retrieving a plurality of coefficients;    multiplying said separate counts by selected ones of said plurality of coefficients and summing the products thereof.    
     
     
         12 . A computer implemented drug discovery system comprising: 
 a PSA estimation module for estimating molecular PSA without reference to three dimensional molecular structures;    a logP estimation module for estimating molecular logP without reference to three dimensional molecular structures;    a molecule selection module for categorizing molecules based on the results of PSA and logP estimation.    
     
     
         13 . The system of  claim 12 , wherein said categorizing comprises separating said molecules into a first class predicted to be readily intestinally absorbed, and a second class predicted to be poorly intestinally absorbed.  
     
     
         14 . The system of  claim 12 , wherein said categorizing comprises ordering at least some of said molecules according to their probabilities of having absorption characteristics similar to known well absorbed compounds.  
     
     
         15 . The system of  claim 12 , wherein said molecule selection module calculates a distance in a PSA-logP plane between molecules being selected and a pre-defined point of said PSA-logP plane.  
     
     
         16 . The system of  claim 12 , additionally comprising a logBB estimation module configured to numerically predict logBB from PSA and logP estimations.  
     
     
         17 . The system of  claim 16 , wherein said categorizing comprises a prioritization of compounds for synthesis and screening.  
     
     
         18 . The system of  claim 12 , wherein said categorizing comprises a prioritization of compounds for synthesis and screening.  
     
     
         19 . A system for drug discovery comprising: 
 a database storing information regarding the atomic constituents and interatomic bonds for a plurality of molecules; and    means for estimating the PSA for each of said plurality of molecules without reference to an energy minimized three dimensional structure for any of said plurality of molecules.    
     
     
         20 . The system of  claim 19 , additionally comprising means for estimating logP for each of said plurality of molecules without reference to an energy minimized three dimensional structure for any of said plurality of molecules.  
     
     
         21 . The system of  claim 20 , additionally comprising means for predicting propensity for intestinal absorption for each of said molecules.