Systems and methods for constructing genomic-based phenotypic models

• US 20030233218A1
• Filed: 06/14/2002
• Published: 12/18/2003
• Est. Priority Date: 06/14/2002
• Status: Active Grant

• Current Assignee: Genomatica Incorporated
• Original Assignee: Genomatica Incorporated
• Inventors: Schilling, Christophe H.
• Granted Patent: US 7,856,317 B2
• Time in Patent Office: Days
• US Class Current: 703/11
• CPC Class Codes: G16B 5/00 ICT specially adapted for modelling or simulations in systems biology, e.g. gene-regulatory networks, protein interaction networks or metabolic networks
• CPC Class Codes: G16B 5/10 Boolean models
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

The invention provides a computer implemented process for constructing a scalable output network model of a bioparticle. The process includes computer implemented steps of: (a) accessing a database of network gene components including an annotated network set of open reading frames (ORFs) of a bioparticle genome; (b) forming a data structure associating the network gene components with network reaction components, the data structure establishing a data set specifying a network model of connectivity and flow of the network reaction components, and (c) transforming the data set into a mathematical description of reactant fluxes defining the network model of connectivity and flow, wherein the mathematical description defines a scalable output network model of a bioparticle.

Claims

1. A computer implemented process for constructing a scalable output network model of a bioparticle, comprising the computer implemented steps of:

(a) accessing a database of network gene components comprising an annotated network set of open reading frames (ORFs) of a bioparticle genome;

(b) forming a data structure associating said network gene components with network reaction components, said data structure establishing a data set specifying a network model of connectivity and flow of said network reaction components, and

(c) transforming said data set into a mathematical description of reactant fluxes defining said network model of connectivity and flow, wherein said mathematical description defines a scalable output network model of a bioparticle.

2. The process of claim 1, wherein forming said data structure further comprises:

(a) selecting an ORF from said annotated network set encoding a gene product having a network reaction function;

(b) forming a data structure comprising said selected gene product, said data structure associating network gene components and network reaction components comprising cognate ORFs, encoded gene products, network reactions and reaction constituents, and

(c) repeating steps (a) and (b) selecting another ORF from said annotated network set until substantially all of said network gene components of said annotated network set have been surveyed for encoding a gene product having a network reaction function to produce a data structure establishing a data set specifying a network model of connectivity and flow.

3. The process of claim 2, further comprising:

(a) determining the occurrence of a constituent gene product for said selected encoded gene product;

(b) determining the occurrence of an additional gene product participating in said network reaction;

(c) determining the occurrence of an alternative network reaction exhibited by a surveyed gene product;

(d) incorporating identified constituent gene products, participating gene products or alternative network reaction into said data structure.

4. The process of claim 1, further comprising incorporating a network reaction that is not gene-encoded and corresponding reaction constituents into said data structure.

5. The process of claim 1, further comprising elemental balancing on at least one network reaction.

6. The process of claim 1, further comprising charge balancing on at least one network reaction.

7. The process of claim 1, further comprising incorporating an exchange reaction for an external reaction component and corresponding reaction constituents into said data structure.

8. The process of claim 7, wherein said external reaction component comprises a metabolite or a biochemical demand constituent.

9. The process of claim 8, wherein said biochemical demand further comprises an aggregate reactant demand flux defining a phenotypic output for growth.

10. The process of claim 9, wherein said phenotypic output for growth comprises biomass production.

11. The process of claim 8, wherein said biochemical demand further comprises an aggregate reactant demand flux defining a phenotypic output selected from the group consisting of energy production, redox equivalent production, catabolite production, biomass precursors, polypeptide production, amino acid production, purine production, pyrimidine production, lipid production, fatty acid production, cofactor production, production of a cell wall component and transport of a metabolite.

12. The process of claim 1, wherein said data structure comprises reactants, products and stoichiometric coefficients.

13. The process of claim 1, wherein said mathematical description comprises linear equations and inequalities.

14. The process of claim 13, wherein said mathematical description comprises a stoichiometric matrix.

15. The process of claim 13, wherein said mathematical description comprises differential equations.

16. The process of claim 1, further comprising calculating a phenotypic output of said network model from said mathematical description.

17. A computer implemented process for constructing a scalable phenotypic output network model, comprising the computer implemented steps of:

(a) accessing a database of network gene components comprising an annotated network set of open reading frames (ORFs) of a bioparticle genome;

(b) forming a data structure associating said network gene components with network reaction components, said data structure establishing a data set specifying a network model of connectivity and flow of said network reaction components;

(c) modifying said data set to enumerate a biochemical demand on said specified network model, and

(d) transforming said modified data set into a mathematical description of reactant fluxes defining said network model of connectivity and flow, wherein said enumerated biochemical demand corresponds to an aggregate reactant demand flux defining a phenotypic output of said network model of a bioparticle.

18. The process of claim 17, wherein forming said data structure further comprises:

(a) selecting an ORF from said annotated network set encoding a gene product having a network reaction function;

(b) forming a data structure comprising said selected gene product, said data structure associating network gene components and network reaction components comprising cognate ORFs, encoded gene products, network reactions and reaction constituents, and

(c) repeating steps (a) and (b) selecting another ORF from said annotated network set until substantially all of said network gene components of said annotated network set have been surveyed for encoding a gene product having a network reaction function to produce a data structure establishing a data set specifying a network model of connectivity and flow.

19. The process of claim 18, further comprising:

(a) determining the occurrence of a constituent gene product for said selected encoded gene product;

(b) determining the occurrence of an additional gene product participating in said network reaction;

(c) determining the occurrence of an alternative network reaction exhibited by a surveyed gene product, and

(d) incorporating identified constituent gene products, participating gene products or alternative network reaction into said data structure.

20. The process of claim 17, further comprising incorporating a network reaction that is not gene-encoded and corresponding reaction constituents into said data structure.

21. The process of claim 17, further comprising elemental balancing on at least one network reaction.

22. The process of claim 17, further comprising charge balancing on at least one network reaction.

23. The process of claim 17, further comprising incorporating an exchange reaction for an external reaction component and corresponding reaction constituents into said data structure.

24. The process of claim 23, wherein said external reaction component comprises a metabolite or a biochemical demand constituent.

25. The process of claim 17, wherein said biochemical demand further comprises an aggregate reactant demand flux defining a phenotypic output.

26. The process of claim 25, wherein said phenotypic output further comprises an aggregate reactant demand flux defining growth.

27. The process of claim 25, wherein said phenotypic output further comprises biomass production.

28. The process of claim 17, wherein said biochemical demand further comprises an aggregate reactant demand flux defining a phenotypic output selected from the group consisting of energy production, redox equivalent production, catabolite production, biomass precursors, polypeptide production, amino acid production, purine production, pyrimidine production, lipid production, fatty acid production, cofactor production, production of a cell wall component and transport of a metabolite.

29. The process of claim 17, wherein said data structure comprises reactants, products and stoichiometric coefficients.

30. The process of claim 17, wherein said mathematical description comprises linear equations and inequalities.

31. The process of claim 30, wherein said mathematical description comprises a stoichiometric matrix.

32. The process of claim 30, wherein said mathematical description comprises differential equations.

33. The process of claim 17, further comprising calculating a phenotypic output of said network model from said mathematical description.

34. A computer implemented process for self-optimizing a network model of a bioparticle, comprising the computer implemented steps:

(a) accessing a database of network gene components comprising an annotated network set of open reading frames (ORFs) of a bioparticle genome;

(b) forming a data structure associating said network gene components with network reaction components, said data structure establishing a data set specifying a network model of connectivity and flow of said network reaction components;

(c) transforming said data set into a mathematical description of reactant fluxes defining said network model of connectivity and flow;

(d) determining the competence of said connectivity and flow within said network model, said competence indicating underinclusion or overinclusion of network reaction component content of said network model, and

(e) identifying an ameliorating network reaction component capable of augmenting said competence of said network model, incorporation of said ameliorating network reaction component into said data structure producing a modified data structure specifying in an optimized network model of said bioparticle.

35. The process of claim 34, wherein said network comprises a metabolic network.

36. The process of claim 35, wherein said metabolic network further comprises a plurality of network pathways of a bioparticle genome.

37. The process of claim 34, wherein forming said data structure further comprises:

(a) selecting an ORF from said annotated network set encoding a gene product having a network reaction function;

(b) forming a data structure comprising said selected gene product, said data structure associating network gene components and network reaction components comprising cognate ORFs, encoded gene products, network reactions and reaction constituents, and

(c) repeating steps (a) and (b) selecting another ORF from said annotated network set until substantially all of said network gene components of said annotated network set have been surveyed for encoding a gene product having a network reaction function to produce a data structure establishing a data set specifying a network model of connectivity and flow.

38. The process of claim 37, further comprising:

(a) determining the occurrence of a constituent gene product for said selected encoded gene product;

(b) determining the occurrence of an additional gene product participating in said network reaction;

(c) determining the occurrence of an alternative network reaction exhibited by a surveyed gene product, and

(d) incorporating identified constituent gene products, participating gene products or alternative network reaction into said data structure.

39. The process of claim 34, further comprising incorporating a network reaction that is not gene-encoded and corresponding reaction constituents into said data structure.

40. The process of claim 34, further comprising elemental balancing on at least one network reaction.

41. The process of claim 34, further comprising charge balancing on at least one network reaction.

42. The process of claim 34, further comprising incorporating an exchange reaction for an external reaction component and corresponding reaction constituents into said data structure.

43. The process of claim 34, further comprising incorporating a biochemical demand into said data structure.

44. The process of claim 34, further comprising:

(a) determining the occurrence of a network reaction component satisfying a macro requirement deficiency in structural architecture of said network model, and

(b) incorporating an identified network reaction component satisfying said macro requirement deficiency into said data structure to supplement said connectivity and flow of said network model.

45. The process of claim 34, further comprising executing a heuristic logic decision algorithm determining confidence of said network reaction components within said data structure.

46. The process of claim 34, wherein said mathematical description comprises linear equations and inequalities.

47. The process of claim 46, wherein said mathematical description comprises a stoichiometric matrix.

48. The process of claim 46, wherein said mathematical description comprises differential equations.

49. The process of claim 34, further comprising determining said competence by solving said mathematical description for a single optimization solution, wherein the ability of said network model to produce a pathway flux indicates a competent network reaction component content.

50. The process of claim 49, further comprising solving said mathematical description for a plurality of single optimization solutions.

51. A computer implemented process for constructing a data structure specifying a network model of a bioparticle, comprising the computer implemented steps:

(a) accessing a database of network gene components comprising an annotated network set of open reading frames (ORFs) of a bioparticle genome;

(b) selecting an ORF from said annotated network set encoding a gene product having a network reaction function;

(c) determining the occurrence of a constituent gene product for said selected encoded gene product;

(d) determining the occurrence of an additional gene product participating in said network reaction;

(e) forming a data structure from said selected and determined gene products, said data structure associating said network gene components and network reaction components comprising cognate ORFs, encoded gene products, network reactions and reaction constituents, and

(f) repeating steps (a)-(e) selecting another ORF from said annotated network set until substantially all of said network gene components of said annotated network set have been surveyed for encoding a gene product having a network reaction function to produce a data structure establishing a data set specifying a network model of connectivity and flow.

52. The process of claim 51, further comprising the steps of:

(a) determining the occurrence of an alternative network reaction exhibited by a surveyed gene product, and

(b) incorporating an identified alternative network reaction and corresponding reaction constituents into said data structure.

53. The process of claim 52, further comprising:

(a) determining the occurrence of a constituent gene product or a gene product participating in said alternative network reaction, and

(b) incorporating an identified constituent gene product or gene product participating in said alternative network reaction into said data structure.

54. The process of claim 51, further comprising incorporating a network reaction that is not gene-encoded and corresponding reaction constituents into said data structure.

55. The process of claim 51, further comprising elemental balancing on at least one network reaction.

56. The process of claim 51, further comprising charge balancing on at least one network reaction.

57. The process of claim 51, further comprising incorporating an exchange reaction for an external reaction component and corresponding reaction constituents into said data structure.

58. The process of claim 57, wherein said external reaction component comprises a metabolite or a biochemical demand constituent.

59. The process of claim 51, further comprising incorporating a biochemical demand into said data structure.

60. The process of claim 59, wherein said biochemical demand further comprises an aggregate reactant demand flux defining a phenotypic output of said network model.

61. The process of claim 51, further comprising:

(a) determining the occurrence of a network reaction component satisfying a macro requirement deficiency in structural architecture of said network model, and

(b) incorporating an identified network reaction component satisfying said macro requirement deficiency into said data structure to supplement said connectivity and flow of said network model.

62. The process of claim 61, wherein said macro requirement deficiency comprises a pathway gap or a pathway dead-end.

63. The process of claim 62, further comprising identifying a singleton reactant.

64. The process of claim 62, further comprising identifying a reactant participating solely in two or more irreversible network reactions.

65. The process of claim 61, wherein said network reaction component comprises a substrate or a product.

66. The process of claim 51, further comprising executing a heuristic logic decision algorithm determining confidence of said network reaction components within said data structure.

67. The process of claim 66, wherein said inclusion of a network reaction component further comprises determining a confidence level from a hierarchical classification.

68. The process of claim 67, wherein said hierarchical classifications are selected from the group consisting of biochemical, genetic, genomic, physiological and simulation modeling data.

69. The process of claim 51, further comprising transforming said data set into a mathematical description of reactant fluxes defining said network model of connectivity and flow of network reaction components.

70. The process of claim 69, wherein said mathematical description comprises linear equations and inequalities.

71. The process of claim 69, wherein said mathematical description comprises a stoichiometric matrix.

72. The process of claim 69, wherein said mathematical description comprises differential equations.

73. The process of claim 51, further comprising performing a validation test.

74. A system for constructing a scalable output network model of a bioparticle, comprising:

(a) an input data set of network gene components comprising an annotated network set of open reading frames (ORFS) of a bioparticle genome;

(b) executable instructions forming a data structure associating said network gene components with network reaction components, said data structure establishing a data set specifying a network model of connectivity and flow of said network reaction components;

(c) executable instructions determining the occurrence of a reaction component satisfying a macro requirement deficiency in structural architecture of said network model, inclusion of an identified reaction component satisfying said macro requirement deficiency in said data structure supplementing said connectivity and flow of said network model;

(d) a heuristic logic decision algorithm determining confidence of said network reaction components within said data structure, and

(e) executable instructions mathematically describing from said data set reactant fluxes defining said network model of connectivity and flow, wherein said mathematical description defines a scalable output network model of a bioparticle.

75. A system for constructing a scalable phenotypic output network model of a bioparticle, comprising:

(a) an input data set of network gene components comprising an annotated network set of open reading frames (ORFs) of a bioparticle genome;

(b) executable instructions forming a data structure associating said network gene components with network reaction components, said data structure establishing a data set specifying a network model of connectivity and flow of said network reaction components;

(c) executable instructions modifying said data set to enumerate a biochemical demand on said specified network model, and

(d) executable instructions mathematically describing from said modified data set reactant fluxes defining said network model of connectivity and flow, wherein said enumerated biochemical demand corresponds to an aggregate reactant demand flux defining a phenotypic output of said network model of said bioparticle.

76. A system for constructing a self-optimizing network model of a bioparticle, comprising:

(a) an input data set of network gene components comprising an annotated network set of open reading frames (ORFs) of a bioparticle genome;

(b) executable instructions forming a data structure associating said network gene components with network reaction components, said data structure establishing a data set specifying a network model of connectivity and flow of said network reaction components;

(c) executable instructions mathematically describing from said data set reactant fluxes defining said network model of connectivity and flow;

(d) executable instructions computing competence of said connectivity and flow within said network model, said competence indicating underinclusion or overinclusion of network reaction component content of said network model, and

(e) executable instructions augmenting said competence of said connectivity and flow within said network model, said executable instructions specifying inclusion or exclusion of an ameliorating network reaction component, wherein incorporation of said ameliorating network reaction component into said data structure produces a modified data structure specifying an optimized network model of said bioparticle.