Method and system for planning, performing, and assessing high-throughput screening of multicomponent chemical compositions and solid forms of compounds

• US 20020177167A1
• Filed: 03/22/2002
• Published: 11/28/2002
• Est. Priority Date: 01/07/2000
• Status: Abandoned Application

• Current Assignee: Donovan Chin, Douglas A. Levinson
• Original Assignee: Donovan Chin, Douglas A. Levinson
• Inventors: Chin, Donovan, Levinson, Douglas A.
• Application Number:
  US10/103,983
• Publication Number:
  US 20020177167A1
• Time in Patent Office: Days
• US Class Current: 435/7.1
• CPC Class Codes: G16B 35/00 ICT specially adapted for in silico combinatorial libraries of nucleic acids, proteins or peptides
• CPC Class Codes: G16C 20/60 In silico combinatorial chemistry
• CPC Class Codes: G16C 20/64 Screening of libraries
• CPC Class Codes: G16C 20/70 Machine learning, data mining or chemometrics
• CPC Class Codes: G16C 20/90 Programming languages; Computing architectures; Database systems; Data warehousing

Abstract

The present application is directed to the use of computerized data processing to plan, perform, and assess the results of high-throughput screening of multicomponent chemical compositions and solid forms of compounds. Systems utilized include databases of molecular descriptors and related compounds and their properties as determined empirically and through simulation, along with multidimensional visualization tools. Methods include methods for determining chemical compositions by performing steps including selecting a plurality of combinations of values of experimental parameters that can be varied by an automated experiment apparatus, determining a set of experimental results, and determining a second plurality of combinations of values based on the set of experimental results. Additional methods include selecting values of parameters that produce a composition, the values being relatively far from areas of rapid change or boundaries between solid forms.

Claims

1. A method for determining one or more multicomponent chemical compositions, comprising the steps of:

selecting a combination of experimental parameters that may be varied by a high-throughput automated experimentation apparatus;

determining a first plurality of distinct combinations of values of the experimental parameters, each combination corresponding to a distinct experiment;

causing the automated experimentation apparatus to conduct a first set of experiments for each of at least a portion of the first plurality of distinct combinations of values of the experimental parameters;

determining a first collection of experimental results of the first set of experiments, the first collection comprising a plurality of individual result sets, each individual result set corresponding to a distinct experiment;

based on the first collection of experimental results, determining a second plurality of distinct combinations of values of the experimental parameters, each combination corresponding to a distinct experiment;

causing the automated experimentation apparatus to conduct a second set of experiments for each of at least a portion of the second plurality of distinct combinations of values of the experimental parameters;

determining a second collection of experimental results of the second set of experiments, the second collection comprising a plurality of individual result sets, each individual result set corresponding to a distinct experiment;

selecting one or more multicomponent chemical compositions of matter based on the first collection of experimental results and the second collection of experimental results.

3. The method of claim 1, wherein the second plurality of distinct combinations of values of experimental parameters is determined based on a comparison of differences of results corresponding to experiments from the first set of experiments with differences of experimental parameters corresponding to the same experiments.

4. The method of claim 3 wherein the comparison comprises a measure of change of experimental results relative to at least one experimental parameter.

5. The method of claim 4 wherein at least a portion of the second plurality of values of the experimental parameters are selected from a range of experimental parameters including values corresponding to a first subset of experiments from the first set of experiments for which the measure of change of experimental results relative to at least one experimental parameter is higher than the measure of change for a second subset of experiments from the first set of experiments.

6. The method of claim 4 wherein at least a portion of the second plurality of values of the experimental parameters are selected from a range of experimental parameters including values corresponding to a first subset of experiments from the first set of experiments for which the measure of change of experimental results relative to at least one experimental parameter is lower than the measure of change for a second subset of experiments from the first set of experiments.

7. The method of claim 1 or 2 wherein at least a portion of the second plurality of distinct combinations of values of experimental parameters is determined based on a comparison of individual results corresponding to experiments from the first set of experiments with one or more target values.

8. The method of claim 1 or 2 wherein at least a portion of the second plurality of distinct combinations of values of experimental parameters is determined based on at least one database query derived from one or more experimental results from the first collection of experimental results.

9. The method of claim 1 or 2, wherein at least a portion of the second plurality of distinct combinations of values of experimental parameters is determined based on the output of at least one simulation that used as input at least one experimental result from the first collection of experimental results.

10. The method of claim 8 wherein the query comprises one or more molecular descriptors characteristic of at least one experiment selected from the first set of experiments.

17. A multicomponent chemical composition of matter determined using the method of claim 1, 2, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15, or 16.

33. The method of claim 1, 11, 13, 18, or 24 wherein the multicomponent chemical composition comprises one or more of:

a pharmaceutical, a nutraceutical, a dietary supplement, an alternative medicine, a sensory material, an agrochemical, a consumer product formulation, an industrial product formulation.

35. The method of claim 1 or 2 wherein the first collection of experimental results is determined using at least one of the techniques selected from the group consisting of mass spectroscopy, HPLC, UV Spectroscopy, fluorescence spectroscopy, gas chromatography, optical density, colorimetry

2. A method for determining one or more solid forms of a compound, comprising the steps of:

selecting a combination of experimental parameters that may be varied by a high-throughput automated experimentation apparatus;

determining a first plurality of distinct combinations of values of the experimental parameters, each combination corresponding to a distinct experiment;

causing the automated experimentation apparatus to conduct a first set of experiments for each of at least a portion of the first plurality of distinct combinations of values of the experimental parameters;

determining a first collection of experimental results of the first set of experiments, the first collection comprising a plurality of individual result sets, each individual result set corresponding to a distinct experiment;

based on the first collection of experimental results, determining a second plurality of distinct combinations of values of the experimental parameters, each combination corresponding to a distinct experiment;

causing the automated experimentation apparatus to conduct a second set of experiments for each of at least a portion of the second plurality of distinct combinations of values of the experimental parameters;

determining a second collection of experimental results of the second set of experiments, the second collection comprising a plurality of individual result sets, each individual result set corresponding to a distinct experiment;

selecting one or more solid forms based on the first collection of experimental results and the second collection of experimental results.

16. The method of claim 2 wherein the second plurality of distinct combinations of values of experimental parameters is determined based on hydrogen-bond-biased simulated annealing monte carlo screening.

11. A method of estimating one or more properties of a multicomponent chemical composition comprising the steps of:

receiving signals representing an experimental result set for each of a plurality of experiments conducted by a high-throughput automated experimentation apparatus;

for each of at least a portion of the experiments, generating a predictive model based on signals characterizing each experimental result set according to the property to be estimated and signals characterizing the experiment with respect to a set of molecular descriptors;

estimating the property for a multicomponent chemical composition by providing signals characterizing the multicomponent chemical composition with respect to the molecular descriptors as input to the predictive model.

15. A method of determining a multicomponent chemical composition of matter, comprising the method of claim 11, 12, 13 or 14, and further comprising the steps of:

based on the estimation, determining a plurality of distinct combinations of values of the experimental parameters, each combination corresponding to a distinct experiment;

causing the automated experimentation apparatus to conduct a second plurality of experiments corresponding to the plurality of distinct combinations of values of the experimental parameters;

receiving signals representing an experimental result set for each of the second plurality of experiments;

selecting a multicomponent chemical composition of matter based on the experimental result sets.

19. The method of claim 18, wherein the discriminator algorithm comprises a predictive model.

20. The method of claim 19, wherein the predictive model further comprises a Kohonen neural network.

12. A method of estimating a property of a solid form of a compound, comprising the steps of:

receiving signals representing experimental result sets for a plurality of experiments conducted by a high-throughput automated experimentation apparatus;

generating a predictive model based on signals characterizing at least a portion of the experimental result sets according to the property to be estimated and signals characterizing the experiments with respect to a set of molecular descriptors;

estimating the property for a solid form of a compound by providing signals characterizing the solid form of the compound with respect to the molecular descriptors as input to the predictive model.

34. The method of claim 2, 12, 14, 21, 25, 30 wherein the compound comprises one or more of:

a pharmaceutical, a nutraceutical, a dietary supplement, an alternative medicine, a sensory material, an agrochemical, a consumer product formulation, an industrial product formulation.

13. A method of estimating a property of a multicomponent chemical composition comprising the steps of:

receiving signals representing a simulation result set for each of a plurality of simulations of a compound;

for at least a portion of the simulations, generating a predictive model based on signals characterizing each simulation result set according to the property to be estimated and signals characterizing the simulation results sets with respect to a set of molecular descriptors;

receiving signals representing experimental result sets for a plurality of experiments conducted by a high-throughput automated experimentation apparatus;

estimating the property for the multicomponent chemical composition by providing signals characterizing the multicomponent chemical composition with respect to the set of molecular descriptors as input to the predictive model.

14. A method of estimating a property of a solid form of a compound comprising the steps of:

receiving signals representing simulation result sets for a plurality of simulations of a compound;

for at least a portion of the simulations, generating a predictive model based on signals characterizing the simulation result sets according to the property to be estimated and signals characterizing the simulation result sets with respect to a set of molecular descriptors;

receiving signals representing experimental result sets for a plurality of experiments conducted by a high-throughput automated experimentation apparatus;

estimating the property for the solid form of the compound by providing signals characterizing the solid form of the compound with respect to the set of molecular descriptors as input to the predictive model.

18. A method of determining one or more multicomponent chemical compositions comprising:

conducting a plurality of experiments using a high-throughput automated experimentation apparatus;

for each experiment, electronically storing;

a set of experimental parameters;

a set of experimental results;

a set of molecular descriptors characterizing an aspect of the experiment;

associating data representing each experiment with previously stored data by querying a database comprising information not derived from the plurality of experiments;

processing at least a portion of the experiment data and the associated previously stored data with a processor programmed to apply a discriminator algorithm to associate at east one experiment with at least one classification;

determining one or more multicomponent chemical compositions based on the at east one classification.

21. A method of determining one or more solid forms of a compound comprising:

conducting a plurality of experiments using a high-throughput automated experimentation apparatus;

for each experiment, electronically storing;

a set of experimental parameters;

a set of experimental results;

a set of molecular descriptors characterizing an aspect of the experiment;

associating data representing each experiment with previously stored data by querying a database comprising information not derived from the plurality of experiments;

processing at least a portion of the experiment data and the associated previously stored data with a processor programmed to apply a discriminator algorithm to associate at least one experiment with at least one classification;

determining one or more solid forms based on the at least one classification.

22. The method of claim 21, wherein the discriminator algorithm comprises a predictive model.

23. The method of claim 22, wherein the predictive model further comprises a Kohonen neural network.

26. The method of claim 24 or 25 wherein the molecular descriptors comprise two or more of the following:

the number of oxygen atoms;

the number of nitrogen atoms;

the number of free carboxylic acid groups;

the number of free primary amine groups;

the number of secondary amine groups;

the total number of amine groups;

the number of hydroxyl groups;

the number of methyl groups;

the number of amide groups;

the number of acid halide groups;

the number of aldehyde groups;

the number of amine oxide groups;

the number of benzene rings;

the number of azo groups;

the number of epoxy rings;

the number of cyclohexyl rings;

the number of isocyanate moities;

the number of ketone moities;

the number of isopropyl groups

the number of ethyl groups;

the number of propyl groups;

the number of butyl groups;

the number of pentyl groups;

the number of hexyl groups;

the number of heptyl groups;

the number of octyl groups;

the number of nonyl groups;

the number of glucose moities;

the number of sucrose moieties;

the number of fructose moities;

the number of amino acids;

the number of peptides;

the molal volume;

the diffusivity;

the molecular weight;

the number of carbon atoms;

the number of halogen atoms;

the total number of N and O atoms;

the proximity effect of N and O;

the number of unsaturated bonds;

the number of aromatic polar substituents;

the critical micelle concentration;

dissociation constant;

partition coefficient;

interatomic distance;

unit cell dimension;

unit cell angle;

pH;

van der Waals radius;

partial charge;

melting point;

boiling point;

sublimation point;

glass transition temperature;

dipole moment;

force constant;

torsional barrier;

inversion barrier;

bond strength;

bond angle;

quantum yield;

delocalization energy;

resonance energy;

compression energy;

molarity;

molality;

density;

viscosity;

dielectric constant;

refractive index;

vapor pressure;

transition energy;

solvent ionizing power;

solubility parameter;

water solubility;

thermal conductivity;

electrical conductivity;

pKa;

atomic radius;

valence;

electronegativity;

electron affinity;

ionization potential;

atomic weight;

atomic number;

stretching force constant;

bending force constant;

volume of activation;

Hammett substituent constant;

chemical shift;

polarizability factor;

NMR coupling constant;

absorbance;

transmittance;

optical purity;

specific rotation;

mole fraction;

mass-to-charge ratio;

charge density.

27. The system of claim 24 or 25, wherein the biological parameters comprise an indicator of one or more of:

permeability of a mammal'"'"'s gastrointestinal membrane;

bioavailability;

taste;

toxicity;

metabolic profile;

color;

smell;

potency;

therapeutic effect.

29. The method of claim 28 wherein the crystal structure of the target compound is predicted using hydrogen-bond-biased simulated annealing.

24. A system for determining a multicomponent chemical composition comprising:

a database comprising at least one table, the at least one table further comprising;

a plurality of molecular descriptors;

a plurality of compound identifiers;

a plurality of compound/descriptor relations associating at least a portion of the compound identifiers with molecular descriptors;

a plurality of empirically determined physical, chemical and biological parameters;

a plurality of compound/parameter relations associating one or more compound identifiers with one or more of the empirically determined physical, chemical and biological parameters;

data representing results from a plurality of experiments performed with a

high-throughput automated experimentation apparatus;

a query system for selecting subsets of related information from the at least one table;

a multidimensional representation generation module capable of generating visual representations of data sets having at least four dimensions;

a plurality of modeling modules, each module capable of receiving information selected by the query system and estimating at least one property of a formulation.

25. A system for determining a solid form of a compound comprising:

a database comprising at least one table, the at least one table further comprising;

a plurality of molecular descriptors;

a plurality of compound identifiers;

a plurality of compound/descriptor relations associating compound identifiers with molecular descriptors;

a plurality of empirically determined physical, chemical and biological parameters;

a plurality of compound/parameter relations associating one or more compound identifiers with one or more of the empirically determined physical, chemical and biological parameters;

data representing results from a plurality of experiments performed with a

high-throughput automated experimentation apparatus;

a query system for selecting subsets of related information from the at least one table;

a multidimensional representation generation module capable of generating visual representations of data sets having at least four dimensions;

a plurality of modeling modules, each module capable of receiving information selected by the query system and estimating at least one property of a formulation.

28. A method for producing crystals comprising:

(a) electronically calculating a set of predicted crystal polymorphs of a target compound;

(b) electronically calculating expected experimental results for at least a portion of the predicted polymorphs;

(c) conducting a first plurality of crystallization experiments using a high-throughput automated experimentation apparatus;

(d) electronically comparing at least a portion of the expected experimental results with the actual experimental results to determine which of the at least a portion of the predicted polymorphs were produced.

30. A method for determining a solid form of a compound comprising:

(a) predicting a crystal structure of a target chemical species;

(b) selecting a first range of conditions for crystal generation;

(c) conducting a first plurality of experiments within the first range of conditions using a high-throughput automated experimentation apparatus;

(d) testing at least a portion of the experimental results for the presence of crystals;

(e) classifying at least a portion of the experiments based on predicted crystal forms;

(f) selecting a second range of conditions for crystal generation based on one or more of the classifications;

(g) conducting a second plurality of experiments within the second range of conditions using the high-throughput automated experimentation apparatus.

31. The method of claim 30 wherein the second range of conditions is determined based on conditions that produced desired crystals.

32. A method for preparing a crystal comprising:

(a) performing simulated hydrogen-bond-biased simulated annealing to predict a plurality of polymorphs of a target compound;

(b) calculating expected properties of at least a portion of the predicted polymorphs;

(c) conducting a plurality of crystallization experiments using a high-throughput automated experimentation apparatus;

(e) comparing measured properties of crystals produced by the plurality of crystallization experiments with the expected properties of at least a portion of the predicted polymorphs to determine which of the at least a portion of the predicted polymorphs were produced by the experiments;

(f) generating a predictive model of the relationship between experimental parameters and polymorphs produced;

(g) calculating a set of experimental parameters for a second set of crystallization experiments from the predictive model.