System, method and computer program product for identifying chemical compounds having desired properties

Abstract

An automatic, partially automatic, and/or manual iterative system, method and/or computer program product for generating chemical entities having desired or specified physical, chemical, functional, and/or bioactive properties. The present invention identifies a set of compounds for analysis; collects, acquires or synthesizes the identified compounds; analyzes the compounds to determine one or more physical, chemical and/or bioactive properties (structure-property data); and uses the structure-property data to identify another set of compounds for analysis in the next iteration. An Experiment Planner generates Selection Criteria and/or one or more Objective Functions for use by a Selector. The Selector searches the Compound Library to identify a subset of compounds (a Directed Diversity Library) that maximizes or minimizes the Objective Functions. The compounds listed in the Directed Diversity Library are then collected, acquired or synthesized, and are analyzed to evaluate their properties of interest. In one embodiment, when a compound in a Directed Diversity Library is available in a Chemical Inventory, the compound is retrieved from the Chemical Inventory instead of re-synthesizing the compound. The Analysis Module receives the compounds of the Directed Diversity Library from the Chemical Inventory and/or the Synthesis Module, analyzes the compounds and outputs Structure-Property data. The Structure-Property data is provided to the Experiment Planner and is also stored in the Structure-Property database. The Experiment Planner defines one or more new Selection Criteria and/or Objective Functions for the next iteration of the invention. In one embodiment, a Structure-Property Model Generator generates Structure-Property Models and provides them to the Experiment Planner which uses the Models to generate subsequent Selection Criteria and/or Objective Function.

What is claimed is:  
     
         1 . A method for identifying chemical compounds having desired properties, comprising the steps of: 
 (1) generating a first set of selection criteria based on one or more desired properties;    (2) selecting a first subset of compounds from a library of compounds based on the first set of selection criteria;    (3) analyzing the first subset of compounds; and    (4) determining, responsive to said analysis of step (3), whether any of the compounds in the first subset of compounds has one or more properties that are substantially similar to the one or more desired properties.    
     
     
         2 . The method according to  claim 1 , further comprising the steps of: 
 (5) generating a second set of selection criteria based on the one or more desired properties and based on one or more properties of one or more of the compounds in the first subset of compounds; and    (6) selecting a second subset of compounds from the library of compounds based on the second set of selection criteria;    (7) analyzing the second subset of compounds; and    (8) determining, responsive to said analysis of step (7), whether any of the compounds in the second subset of compounds has one or more properties that are substantially similar to the one or more desired properties.    
     
     
         3 . The method according to  claim 1 , wherein step (1) comprises the steps of: 
 (a) generating one or more structure-property models that predict properties of compounds; and    (b) training the one or more structure-property models to minimize error between predicted properties and actual properties.    
     
     
         4 . The method according to  claim 3 , wherein step (1)(a) comprises the step of: 
 (i) generating at least one neural network structure-property model.    
     
     
         5 . The method according to  claim 3 , wherein step (1)(a) comprises the step of: 
 (i) generating at least one Neuro-Fuzzy structure-property model based on neural networks and fuzzy logic.    
     
     
         6 . The method according to  claim 3 , wherein step (1)(a) comprises the step of: 
 (i) generating at least one generalized regression neural network structure-property model that employs K-nearest-neighbor classifiers.    
     
     
         7 . The method according to  claim 3 , wherein step (1)(b) comprises training the one or more structure-property models using one or more of the following techniques: 
 (i) gradient minimization;    (ii) Monte Carlo;    (iii) simulated annealing;    (iv) evolutionary programming; and    (v) genetic algorithms.    
     
     
         8 . The method according to  claim 1 , wherein step (1) comprises the step of: 
 (a) generating one or more objective functions from the first set of selection criteria, each objective function specifying a collection of selection criteria that a selected compound should exhibit.    
     
     
         9 . The method according to  claim 1 , wherein step (2) comprises the steps of: 
 (a) selecting an initial set of one or more compounds;    (b) assessing the initial set of one or more compounds;    (c) modifying the initial set of one or more compounds to generate a new set of one or more compounds;    (d) assessing the new set of one or more compounds;    (e) replacing the initial set of one or more compounds with the new set of one or more compounds when the new set of one or more compounds is determined to be better than the initial set of one or more compounds; and    (f) repeating steps (1)(a)-(1)(e) a number of times; and    (g) outputting a set of compounds as the first subset of compounds.    
     
     
         10 . The method according to  claim 1 , wherein step (2) comprises selecting a first subset of compounds using one or more of the following techniques: 
 (a) Monte Carlo;    (b) simulated annealing;    (c) evolutionary programming; and    (d) genetic algorithms.