Method and apparatus for automated design of chemical synthesis routes

• US 20030220716A1
• Filed: 05/30/2003
• Published: 11/27/2003
• Est. Priority Date: 03/12/1999
• Status: Abandoned Application

• Current Assignee: Numerate, Inc.
• Original Assignee: PHARMIX CORP. (Numerate, Inc.)
• Inventors: Yu, Jessen, Mydlowec, William
• Application Number:
  US10/452,006
• Publication Number:
  US 20030220716A1
• Time in Patent Office: Days
• US Class Current: 700/268
• CPC Class Codes: B01J 2219/007 Simulation or vitual synthesis
• CPC Class Codes: C40B 40/00 Libraries per se, e.g. arrays, mixtures
• CPC Class Codes: G16C 20/10 Analysis or design of chemical reactions, syntheses or processes

Abstract

Method and apparatus for designing a synthesis route for a target molecule is provided. The method for designing a synthesis route for a target molecule comprises: generating a plurality of individuals, wherein each individual encodes a synthesis route; decoding each individual to produce a synthesis route comprising at least one reactant molecules and at least one reaction; and determining how well the synthesis route satisfies a design goal. A computer readable medium containing instructions for a computer program executable by the computer to perform a method for designing a synthesis route for a target molecule is also provided. The apparatus comprises a parallel computer system for executing instructions of a computer program to perform a method for designing a synthesis route for a target molecule.

Claims

21. A computer readable medium containing instructions for a computer program executable by the computer to perform a method for designing a synthesis route for a target molecule, the method comprising:

generating at least one individual, wherein each individual encodes a synthesis route;

decoding each individual to produce a synthesis route comprising at least one reactant molecule and at least one reaction, and determining whether the synthesis route satisfies a design goal.

22. The computer readable medium of claim 21, wherein the decoding step comprises:

retrieving at least one reactant molecule from a data structure of molecules;

retrieving at least one set of reaction conditions from a data structure of reaction conditions; and

evaluating at least one reaction utilizing the at least one reactant molecule and the at least one set of reaction conditions to output at least one product molecule.

23. The computer readable medium of claim 22, wherein the evaluation is performed utilizing a computer-based reaction predictor.

24. The computer readable medium of claim 21, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a structural similarity between a target molecule and a final product molecule of the synthesis route.

25. The computer readable medium of claim 24, wherein the step of determining structural similarity comprises comparing the final product molecule and the target molecule utilizing a graph isomorphism algorithm.

26. The computer readable medium of claim 24, wherein the step of determining structural similarity comprises generating and comparing reduced representations of the final product molecule and the target molecule.

27. The computer readable medium of claim 21, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a yield value of the final product molecule.

28. The computer readable medium of claim 21, wherein the method further comprises:

performing a mutation operation on at least a part of the at least one individual; and

determining whether the mutated individual satisfies the design goal.

29. An apparatus for designing a synthesis route for a target molecule, comprising:

a parallel computer system for executing instructions of a computer program to perform a method for designing a synthesis route for a target molecule, the method comprising;

generating at least one individual, wherein each individual encodes a synthesis route;

decoding each individual to produce a synthesis route comprising at least one reactant molecule and at least one reaction, and determining how well the synthesis route satisfies a design goal.

30. The apparatus of claim 29, wherein the decoding step comprises:

retrieving at least one reactant molecule from a data structure of molecules;

retrieving at least one set of reaction conditions from a data structure of reaction conditions; and

evaluating at least one reaction utilizing the at least one reactant molecule and the at least one set of reaction conditions to output at least one product molecule.

31. The apparatus of claim 30, wherein the evaluation is performed utilizing a computer-based reaction predictor.

32. The apparatus of claim 29, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a structural similarity between a target molecule and a final product molecule of the synthesis route.

33. The apparatus of claim 29, wherein the step of determining structural similarity comprises comparing the final product molecule and the target molecule utilizing a graph isomorphism algorithm.

34. The apparatus of claim 29, wherein the step of determining structural similarity comprises generating and comparing reduced representations of the final product molecule and the target molecule.

35. The apparatus of claim 34, wherein the reduced representations are selected from a bit string distance metric or a degree of fitness vector with floating point numbers, wherein each floating point number represents a degree of fit.

36. The apparatus of claim 29, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a yield value of the final product molecule.

37. The apparatus of claim 29, wherein the method further comprises:

performing a mutation operation on at least a part of the at least one individual; and

determining whether the mutated individual satisfies the design goal.

38. A method for designing a synthesis route for a target molecule, comprising:

generating at least one individual, wherein the at least one individual is a synthesis route comprising at least one reactant molecule and at least one reaction, and determining whether the synthesis route satisfies a design goal.

39. The method of claim 38, wherein the individual is a plurality of instructions that are executed to produce the synthesis route.

40. The method of claim 39, wherein the individual conforms to a constrained structure.

41. The method of claim 38, wherein the determining step further comprises evaluating at least one reaction utilizing the at least one reactant molecule and the at least one set of reaction conditions to output at least one product molecule.

42. The method of claim 41, wherein the evaluation is performed utilizing a computer-based reaction predictor.

43. The method of claim 41, wherein the evaluation is accomplished by physically performing the reaction.

44. The method of claim 38, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a structural similarity between a target molecule and a final product molecule of the synthesis route.

45. The method of claim 44, wherein the step of determining structural similarity comprises comparing the final product molecule and the target molecule utilizing a graph isomorphism algorithm.

46. The method of claim 44, wherein the step of determining structural similarity comprises generating and comparing reduced representations of the final product molecule and the target molecule.

47. The method of claim 46, wherein the reduced representation comprises a bit string and wherein the step of comparing reduced representations utilizes a bit string distance metric.

48. The method of claim 46, wherein the reduced representation comprises a degree of fitness vector with floating point numbers, wherein each floating point number represents a degree of fit.

49. The method of claim 38, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a yield value of the final product molecule.

50. The method of claim 38, wherein the determining step applies a point search algorithm or a beam search algorithm.

51. The method of claim 38, wherein the algorithm is selected from a hill climbing, greedy, Monte Carlo, simulated annealing, or genetic algorithm.

52. The method of claim 51, wherein the algorithm is simulated annealing.

53. The method of claim 50, wherein the algorithm is a beam search algorithm and further comprising performing a crossover operation, the crossover operation comprising:

combining at least one portion from the at least one individual and at least one portion from another individual to produce a new individual.

54. The method of claim 50, wherein the algorithm is a point-search or beam-search algorithm and further comprising performing a mutation operation, the mutation operation comprising:

replacing at least one portion of the at least one individual with a randomly generated portion to produce a new individual.

55. The method of claim 50, wherein the algorithm is a point-search or beam-search algorithm and further comprising the step of performing a molecule noise operation, the molecule noise operation comprising:

generating a similarity value;

selecting at least one molecule encoded by the at least one individual; and

modifying the at least one individual to encode a new molecule based on the similarity value of the selected at least one molecule.

56. The method of claim 38, wherein the steps are performed utilizing a computer system having a plurality of processors.

57. A computer readable medium containing instructions for a computer program executable by the computer to perform a method for designing a synthesis route for a target molecule, the method comprising:

generating at least one individual, wherein the at least one individual is a synthesis route comprising at least one reactant molecule and at least one reaction, and determining whether the synthesis route satisfies a design goal.

1. A method for designing a synthesis route for a target molecule, comprising:

generating at least one individual, wherein each individual encodes a synthesis route;

decoding each individual to produce a synthesis route comprising at least one reactant molecule and at least one reaction, and determining whether the synthesis route satisfies a design goal.

2. The method of claim 1, wherein the individual encodes a synthesis route as a program comprising a plurality of instructions that are executed to produce the synthesis route.

3. The method of claim 2, wherein the program conforms to a constrained structure.

4. The method of claim 1, wherein the decoding step comprises retrieving at least one reactant molecule from a data structure of molecules.

5. The method of claim 4, wherein the decoding step further comprises retrieving at least one set of reaction conditions from a data structure of reaction conditions.

6. The method of claim 5, wherein the decoding step further comprises evaluating at least one reaction utilizing the at least one reactant molecule and the at least one set of reaction conditions to output at least one product molecule.

7. The method of claim 6, wherein the evaluation is performed utilizing a computer-based reaction predictor.

8. The method of claim 6, wherein the evaluation is accomplished by physically performing the reaction.

9. The method of claim 1, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a structural similarity between a target molecule and a final product molecule of the synthesis route.

10. The method of claim 9, wherein the step of determining structural similarity comprises comparing the final product molecule and the target molecule utilizing a graph isomorphism algorithm.

11. The method of claim 9, wherein the step of determining structural similarity comprises generating and comparing reduced representations of the final product molecule and the target molecule.

12. The method of claim 11, wherein the reduced representation comprises a bit string and wherein the step of comparing reduced representations utilizes a bit string distance metric.

13. The method of claim 11, wherein the reduced representation comprises a degree of fitness vector with floating point numbers, wherein each floating point number represents a degree of fit.

14. The method of claim 1, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a yield value of the final product molecule.

15. The method of claim 1, wherein the determining step applies a point search algorithm.

16. The method of claim 15, wherein the point search algorithm is selected from a hill climbing, greedy, Monte Carlo, or simulated annealing algorithm.

17. The method of claim 16, wherein the point search algorithm is simulated annealing.

18. The method of claim 1, further comprising:

performing a mutation operation on at least a part of the at least one individual; and

determining whether the mutated individual satisfies the design goal.

19. The method of claim 1, wherein the steps are performed utilizing a computer system having a plurality of processors.

20. The method of claim 1, wherein said method is performed at least two times in parallel by a computer system having a plurality of processors.

58. The computer readable medium of claim 57, wherein the determining step comprises retrieving at least one reactant molecule from a data structure of molecules;

retrieving at least one set of reaction conditions from a data structure of reaction conditions; and

evaluating at least one reaction utilizing the at least one reactant molecule and the at least one set of reaction conditions to output at least one product molecule.

59. The computer readable medium of claim 58, wherein the evaluation is performed utilizing a computer-based reaction predictor.

60. The computer readable medium of claim 57, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a structural similarity between a target molecule and a final product molecule of the synthesis route.

62. The computer readable medium of claim 60, wherein the step of determining structural similarity comprises comparing the final product molecule and the target molecule utilizing a graph isomorphism algorithm.

63. The computer readable medium of claim 57, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a yield value of the final product molecule.

64. The computer readable medium of claim 57, wherein the computer executes a hill climbing, greedy, Monte Carlo, simulated annealing or genetic algorithm.

62-1. The computer readable medium of claim 60, wherein the step of determining structural similarity comprises generating and comparing reduced representations of the final product molecule and the target molecule.

65. An apparatus for designing a synthesis route for a target molecule, comprising:

a parallel computer system for executing instructions of a computer program to perform a method for designing a synthesis route for a target molecule, the method comprising;

generating at least one individual, wherein each individual is synthesis route comprising at least one reactant molecules and at least one reaction; and

determining how well the synthesis route satisfies a design goal.

66. The apparatus of claim 65, wherein the determining step comprises:

retrieving at least one reactant molecule from a data structure of molecules;

retrieving at least one set of reaction conditions from a data structure of reaction conditions; and

evaluating at least one reaction utilizing the at least one reactant molecule and the at least one set of reaction conditions to output at least one product molecule.

67. The apparatus of claim 66, wherein the evaluation is performed utilizing a computer-based reaction predictor.

68. The apparatus of claim 65, wherein the step of determining whether the synthesis route satisfies a design goal comprises determining a structural similarity between a target molecule and a final product molecule of the synthesis route.

69. The apparatus of claim 68, wherein the step of determining structural similarity comprises comparing the final product molecule and the target molecule utilizing a graph isomorphism algorithm.

70. The apparatus of claim 68, wherein the step of determining structural similarity comprises generating and comparing reduced representations of the final product molecule and the target molecule.