Models and Languages for Computational Systems Biology Lecture 1

Transcription

1 Models and Languages for Computational Systems Biology Lecture 1 Jane Hillston. LFCS and CSBE, University of Edinburgh 13th January 2011

2 Outline Introduction Motivation Measurement, Observation and Induction Models and Languages Case Study Challenges

3 Systems Biology Biological advances mean that much more is now known about the components of cells and the interactions between them.

4 Systems Biology Biological advances mean that much more is now known about the components of cells and the interactions between them. Systems biology aims to develop a better understanding of the processes involved.

5 Systems Biology Biological advances mean that much more is now known about the components of cells and the interactions between them. Systems biology aims to develop a better understanding of the processes involved. It involves taking a systems theoretic view of biological processes analysing inputs and outputs and the relationships between them.

6 Systems Biology Biological advances mean that much more is now known about the components of cells and the interactions between them. Systems biology aims to develop a better understanding of the processes involved. It involves taking a systems theoretic view of biological processes analysing inputs and outputs and the relationships between them. A radical shift from earlier reductionist approaches, systems biology aims to provide a conceptual basis and a methodology for reasoning about biological phenomena.

7 Relation to Computer Science Formalisms from theoretical computer science have found a new role in developing models for systems biology, allowing biologists to test hypotheses and prioritise experiments.

8 Relation to Computer Science Formalisms from theoretical computer science have found a new role in developing models for systems biology, allowing biologists to test hypotheses and prioritise experiments. Perhaps more than any other discipline computer scientists have a track record of developing languages and formal reasoning techniques to support them.

9 What is Systems Biology? The principal aim of systems biology is to provide both a conceptual basis and working methodologies for the scientific explanation of biological phenomena Olaf Wolkenhauer

10 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

11 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

12 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

13 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

14 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

15 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

16 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

17 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

18 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

19 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

20 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

21 Biological Phenomena As an approach systems biology can be applied to many different biological systems at different scales.

22 Biological Phenomena As an approach systems biology can be applied to many different biological systems at different scales. For example, from gene regulation within the nucleus of a cell, to whole organs, or even complete organisms.

23 Biological Phenomena As an approach systems biology can be applied to many different biological systems at different scales. For example, from gene regulation within the nucleus of a cell, to whole organs, or even complete organisms. The biological phenomena to be studied will clearly depend on the type of system being investigated.

24 Biological Phenomena As an approach systems biology can be applied to many different biological systems at different scales. For example, from gene regulation within the nucleus of a cell, to whole organs, or even complete organisms. The biological phenomena to be studied will clearly depend on the type of system being investigated. A grand challenge for systems biology is to develop multi-scale models which seek to account for high-level behaviour (at the level of the whole organisms) at all levels down to the intra-cellular processes.

25 Biochemical Pathways At the intra-cellular level we can distinguish three distinct types of pathways or networks

26 Biochemical Pathways At the intra-cellular level we can distinguish three distinct types of pathways or networks Gene networks: Genes control the production of proteins but are themselves regulated by the same or different proteins.

27 Biochemical Pathways At the intra-cellular level we can distinguish three distinct types of pathways or networks Gene networks: Genes control the production of proteins but are themselves regulated by the same or different proteins. Signal transduction networks: External stimuli initiate messages that are carried through a cell via a cascade of biochemical reactions.

28 Biochemical Pathways At the intra-cellular level we can distinguish three distinct types of pathways or networks Gene networks: Genes control the production of proteins but are themselves regulated by the same or different proteins. Signal transduction networks: External stimuli initiate messages that are carried through a cell via a cascade of biochemical reactions. Metabolic pathways: The survival of the cell depends on its ability to transform nutrients into energy.

29 Biochemical Pathways At the intra-cellular level we can distinguish three distinct types of pathways or networks Gene networks: Genes control the production of proteins but are themselves regulated by the same or different proteins. Signal transduction networks: External stimuli initiate messages that are carried through a cell via a cascade of biochemical reactions. Metabolic pathways: The survival of the cell depends on its ability to transform nutrients into energy. But these distinctions are to some extent arbitrary as models may include elements of more than one pathway type.

30 Signal transduction pathways All signalling is biochemical:

31 Signal transduction pathways All signalling is biochemical: Increasing protein concentration broadcasts the information about an event.

32 Signal transduction pathways All signalling is biochemical: Increasing protein concentration broadcasts the information about an event. The message is received by a concentration dependent response at the protein signal s site of action.

33 Signal transduction pathways All signalling is biochemical: Increasing protein concentration broadcasts the information about an event. The message is received by a concentration dependent response at the protein signal s site of action. This stimulates a response at the signalling protein s site of action.

34 Signal transduction pathways All signalling is biochemical: Increasing protein concentration broadcasts the information about an event. The message is received by a concentration dependent response at the protein signal s site of action. This stimulates a response at the signalling protein s site of action. Signals propagate through a series of protein accumulations.

35 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

36 Measurement, Observation and Induction Robot Scientist project Kell, King, Muggleton et al.

37 Measurement, Observation and Induction Robot Scientist project Kell, King, Muggleton et al. Combination of machine learning for hypothesis generation and genetic algorithms for automatic experimental tuning.

38 Measurement, Observation and Induction Robot Scientist project Kell, King, Muggleton et al. Combination of machine learning for hypothesis generation and genetic algorithms for automatic experimental tuning. Experiments are carried out by a robot.

39 Measurement, Observation and Induction Robot Scientist project Kell, King, Muggleton et al. Combination of machine learning for hypothesis generation and genetic algorithms for automatic experimental tuning. Experiments are carried out by a robot. Data is generated at rates which exceed what is possible when there are humans in the loop.

40 Measurement, Observation and Induction Robot Scientist project Kell, King, Muggleton et al. Combination of machine learning for hypothesis generation and genetic algorithms for automatic experimental tuning. Experiments are carried out by a robot. Data is generated at rates which exceed what is possible when there are humans in the loop. Moreover the intelligent experiment selection strategy is competitive with (good) human strategies, and significantly outperforms cheapest and random selection strategies.

41 The Robot Scientist Background Knowledge Machine learning Consistent hypotheses Analysis Final Hypothesis Experiment selection Experiments Robot Results

42 The Robot Scientist Background Knowledge Machine learning Consistent hypotheses Analysis Final Hypothesis Experiment selection Experiments Robot Results No human intellectual input in the design of experiments or the interpretation of data.

43 The Robot Scientist Background Knowledge Machine learning Consistent hypotheses Analysis Final Hypothesis Experiment selection Experiments Robot Results No human intellectual input in the design of experiments or the interpretation of data. Integrates scientific discovery software with laboratory robotics.

44 The Robot Scientist The initial robot, Adam, solved a problem related to identifying the gene in yeast responsible for an enzyme crucial to the production of the amino acid lysine. A successor robot, Eve, is now in operation at Aberystwyth University. See press report at news/science/article ece

45 Systems Biology Methodology Natural System Measurement Observation Biological Phenomena Explanation Interpretation Induction Modelling Systems Analysis Deduction Inference Formal System

46 Formal Systems There are two alternative approaches to constructing dynamic models of biochemical pathways commonly used by biologists:

47 Formal Systems There are two alternative approaches to constructing dynamic models of biochemical pathways commonly used by biologists: Ordinary Differential Equations: continuous time, continuous behaviour (concentrations), deterministic.

48 Formal Systems There are two alternative approaches to constructing dynamic models of biochemical pathways commonly used by biologists: Ordinary Differential Equations: continuous time, continuous behaviour (concentrations), deterministic. Stochastic Simulation: continuous time, discrete behaviour (no. of molecules), stochastic.

49 The objective of modelling

50 The objective of modelling The principal aim of systems biology is to provide both a conceptual basis and working methodologies for the scientific explanation of biological phenomena Olaf Wolkenhauer

51 The objective of modelling The principal aim of systems biology is to provide both a conceptual basis and working methodologies for the scientific explanation of biological phenomena Olaf Wolkenhauer The key word here is explanation.

52 The objective of modelling The principal aim of systems biology is to provide both a conceptual basis and working methodologies for the scientific explanation of biological phenomena Olaf Wolkenhauer The key word here is explanation. The role of modelling is to explore possible explanations more abstractly and more efficiently than might be done with wet lab experiments.

53 Case Study: Circadian Rhythms Overview The study by Locke et al. ( Extension of a genetic network model by iterative experimentation and mathematical analysis. Molecular Systems Biology) focuses on the circadian rhythms in plants, combining mathematical models and molecular biology.

54 Case Study: Circadian Rhythms Overview The study by Locke et al. ( Extension of a genetic network model by iterative experimentation and mathematical analysis. Molecular Systems Biology) focuses on the circadian rhythms in plants, combining mathematical models and molecular biology. Their objective is to identify the genes (and proteins) responsible for maintaining the daily rhythms observed in the plants.

55 Case Study: Circadian Rhythms Overview The study by Locke et al. ( Extension of a genetic network model by iterative experimentation and mathematical analysis. Molecular Systems Biology) focuses on the circadian rhythms in plants, combining mathematical models and molecular biology. Their objective is to identify the genes (and proteins) responsible for maintaining the daily rhythms observed in the plants. The research exploits an interplay between mathematical models, experiments in the laboratory and literature search.

56 Case Study: Circadian Rhythms Overview The study by Locke et al. ( Extension of a genetic network model by iterative experimentation and mathematical analysis. Molecular Systems Biology) focuses on the circadian rhythms in plants, combining mathematical models and molecular biology. Their objective is to identify the genes (and proteins) responsible for maintaining the daily rhythms observed in the plants. The research exploits an interplay between mathematical models, experiments in the laboratory and literature search. It is held up as an exemplar of what systems biology is trying to achieve, and the breakthroughs that it can bring about when it is successful.

57 Case Study: Circadian Rhythms Initial Model From initial experiments Locke et al. identified a two genes and two proteins which appeared to operate in a simple loop: LHY TOC1 LHY TOC1 An initial mathematical model (ODEs) was constructed to capture this model.

58 Case Study: Circadian Rhythms Role of Mathematics Initial simulations with the mathematical model showed good agreement with the experimental data for some of the observed phenomena but significant discrepancies for others.

59 Case Study: Circadian Rhythms Role of Mathematics Initial simulations with the mathematical model showed good agreement with the experimental data for some of the observed phenomena but significant discrepancies for others. Experiments were then undertaken with the mathematical model to find an alternative model which was biologically plausible but produced a better fit.

60 Case Study: Circadian Rhythms Role of Mathematics Initial simulations with the mathematical model showed good agreement with the experimental data for some of the observed phenomena but significant discrepancies for others. Experiments were then undertaken with the mathematical model to find an alternative model which was biologically plausible but produced a better fit. These mathematical experiments conjectured a network with two interacting loops.

61 Case Study: Circadian Rhythms Elaborated Model X X LHY LHY Two new genes were introduced to the model which now has interlocking loops and more complex feedback. TOC1 TOC1 Y Y

62 Case Study: Circadian Rhythms Elaborated Model X X LHY LHY Two new genes were introduced to the model which now has interlocking loops and more complex feedback. TOC1 TOC1 Y The simulation results from this model showed much better agreement with the observed data. Y

63 Case Study: Circadian Rhythms Validating the Model The researchers then sought to identify the new genes X and Y. Searching the literature elicited several candidate genes which previous experimental studies had suggested were implicated in the circadian rhythm. In particular, knockout data for one, GIGANTEA (GI), coincided with the pattern from simulation experiments of the original model with a single loop. Subsequent wet lab experiments have reinforced this impression that GI is gene Y.

64 Formal Systems Revisited The case study used mathematics (i.e. a set of ODEs) directly as the formal system.

65 Formal Systems Revisited The case study used mathematics (i.e. a set of ODEs) directly as the formal system. Previous experience in other application domains has shown that there can be benefits to interposing a formal model between the system and the underlying mathematical model.

66 Formal Systems Revisited The case study used mathematics (i.e. a set of ODEs) directly as the formal system. Previous experience in other application domains has shown that there can be benefits to interposing a formal model between the system and the underlying mathematical model. Using such an intermediary can aid is model construction and model understanding, it can reduce errors and provide access to logic-based analysis techniques such as model checking.

67 Formal Systems Revisited The case study used mathematics (i.e. a set of ODEs) directly as the formal system. Previous experience in other application domains has shown that there can be benefits to interposing a formal model between the system and the underlying mathematical model. Using such an intermediary can aid is model construction and model understanding, it can reduce errors and provide access to logic-based analysis techniques such as model checking. Moreover taking this high-level programming style approach offers the possibility of different compilations to different mathematical models.

68 Integrated Analysis System Mathematical Model e.g ODE or SSA

69 Integrated Analysis System Stochastic Process Algebra Model ODE SSA Model Checking

70 Integrated Analysis Stochastic Process Algebra Model ODE SSA Model Checking Analysis

71 Challenges Individual vs. Population Noise vs. Determinism Modularity vs. Infinite Regress Dealing with the Unknown

72 Individual vs. Population behaviour Biochemistry is concerned with the reactions between individual molecules and so it is often more natural to model at this level.

73 Individual vs. Population behaviour Biochemistry is concerned with the reactions between individual molecules and so it is often more natural to model at this level. However experimental data is usually more readily available in terms of populations rather than individual molecules cf. average reaction rates rather than the forces at play on an individual molecule in a particular physical context.

74 Individual vs. Population behaviour Biochemistry is concerned with the reactions between individual molecules and so it is often more natural to model at this level. However experimental data is usually more readily available in terms of populations rather than individual molecules cf. average reaction rates rather than the forces at play on an individual molecule in a particular physical context. These should be regarded as alternatives, each being appropriate for some models. The challenge then becomes when to use which approach.

75 Individual vs. Population behaviour Biochemistry is concerned with the reactions between individual molecules and so it is often more natural to model at this level. However experimental data is usually more readily available in terms of populations rather than individual molecules cf. average reaction rates rather than the forces at play on an individual molecule in a particular physical context. These should be regarded as alternatives, each being appropriate for some models. The challenge then becomes when to use which approach. Note that given a large enough number of molecules an individuals model will (in many circumstances) be indistinguishable from the a population level model.

76 Noise vs. Determinism With perfect knowledge the behaviour of a biochemical reaction would be deterministic.

77 Noise vs. Determinism With perfect knowledge the behaviour of a biochemical reaction would be deterministic. However, in general, we do not have the requisite knowledge of thermodynamic forces, exact relative positions, temperature, velocity etc.

78 Noise vs. Determinism With perfect knowledge the behaviour of a biochemical reaction would be deterministic. However, in general, we do not have the requisite knowledge of thermodynamic forces, exact relative positions, temperature, velocity etc. Thus a reaction appears to display stochastic behaviour.

79 Noise vs. Determinism With perfect knowledge the behaviour of a biochemical reaction would be deterministic. However, in general, we do not have the requisite knowledge of thermodynamic forces, exact relative positions, temperature, velocity etc. Thus a reaction appears to display stochastic behaviour. When a large number of such reactions occur, the randomness of the individual reactions can cancel each other out and the apparent behaviour exhibits less variability.

80 Noise vs. Determinism With perfect knowledge the behaviour of a biochemical reaction would be deterministic. However, in general, we do not have the requisite knowledge of thermodynamic forces, exact relative positions, temperature, velocity etc. Thus a reaction appears to display stochastic behaviour. When a large number of such reactions occur, the randomness of the individual reactions can cancel each other out and the apparent behaviour exhibits less variability. However, in some systems the variability in the stochastic behaviour plays a crucial role in the dynamics of the system.

81 Comparing stochastic simulation and ODEs Consider a Michaelis Menten reaction in which a substrate S is transformed to a product S via a complex C formed with an enzyme E. S b/u C c S E It is relatively straightforward to contrast the results of the two methods. We compare the results of 2000 runs of the stochastic algorithm simulating a system with initial molecular populations S 0 = 100, E 0 = 10, C 0 = 0, S 0 = 0 and a volume of 1000 units.

82 Results for S 0 = 100, E 0 = 10, C 0 = 0, S 0 = 0 (vol 1000)

83 Results for S 0 = 100, E 0 = 10, C 0 = 0, S 0 = 0 (vol 1000)

84 In vivo behaviour compared with model behaviour However, it is worth bearing in mind that an actual in vivo biochemical reaction would follow just one of the many random curves that average together producing the closely fitting mean. This curve may deviate significantly from that of the deterministic approach, and thus call into question its validity.

85 In vivo behaviour compared with model behaviour However, it is worth bearing in mind that an actual in vivo biochemical reaction would follow just one of the many random curves that average together producing the closely fitting mean. This curve may deviate significantly from that of the deterministic approach, and thus call into question its validity. But this does not mean that the randomness exhibited by a particular stochastic simulation trajectory will be the same as the randomness of a particular in vivo reaction. Indeed, a set of stochastic simulation trajectories (ensemble) is usually averaged before any conclusions are drawn.

86 Comparing results at lower population sizes S 0 = 10, E 0 = 1, C 0 = 0, S 0 = 0 (vol 100)

87 Mean results for 11, 110 and 1100 molecules

88 Vliar-Kueh-Barkai-Leibler Circadian clock (cartoon)

89 Circadian clock (deterministically... )

90 Circadian clock (... and stochastically)

91 Modularity vs. Infinite Regress As computer scientists we are firm believers in modularity and compositionality. When it comes to biochemical pathways opinion amongst biologists is divided about whether is makes sense to take a modular view of cellular pathways.

92 Modularity vs. Infinite Regress As computer scientists we are firm believers in modularity and compositionality. When it comes to biochemical pathways opinion amongst biologists is divided about whether is makes sense to take a modular view of cellular pathways. Some biologists (e.g. Leibler) argue that there is modularity, naturally occuring, where they define a module relative to a biological function.

93 Modularity vs. Infinite Regress As computer scientists we are firm believers in modularity and compositionality. When it comes to biochemical pathways opinion amongst biologists is divided about whether is makes sense to take a modular view of cellular pathways. Some biologists (e.g. Leibler) argue that there is modularity, naturally occuring, where they define a module relative to a biological function. Others such as Cornish-Bowden are much more skeptical and cite the problem of infinite regress as being insurmountable.

94 The problem of Infinite Regress

95 The problem of Infinite Regress

96 The problem of Infinite Regress

97 The problem of Infinite Regress

98 The problem of Infinite Regress

99 The problem of Infinite Regress

100 The problem of Infinite Regress

101 The problem of Infinite Regress

102 The problem of Infinite Regress

103 The problem of Infinite Regress

104 The problem of Infinite Regress

105 The problem of Infinite Regress

106 The problem of Infinite Regress

107 Dealing with the Unknown There is a fundamental challenge when modelling cellular pathways that little is known about some aspects of cellular processes.

108 Dealing with the Unknown There is a fundamental challenge when modelling cellular pathways that little is known about some aspects of cellular processes. In some cases this is because no experimental data is available, or that the experimental data that is available is inconsistent.

109 Dealing with the Unknown There is a fundamental challenge when modelling cellular pathways that little is known about some aspects of cellular processes. In some cases this is because no experimental data is available, or that the experimental data that is available is inconsistent. In other cases the data is unknowable because experimental techniques do not yet exist to collect the data, or those that do involve modification to the system.

110 Dealing with the Unknown There is a fundamental challenge when modelling cellular pathways that little is known about some aspects of cellular processes. In some cases this is because no experimental data is available, or that the experimental data that is available is inconsistent. In other cases the data is unknowable because experimental techniques do not yet exist to collect the data, or those that do involve modification to the system. Even when data exists the quality is often very poor.

111 References R.D. King, K.E. Whelan, F.M. Jones, P.J.K. Reiser, C.H. Bryant, S. Muggleton, D.B. Kell, S. Oliver Functional Genomic Hypothesis Generation and Experimentation by a Robot Scientist Nature, 427, pp , J.C.W. Locke, M.M. Southern, L. Kozma-Bognár, V. Hibberd, P.E. Brown, M.S. Turner and A.J. Millar. Extension of a genetic network model by iterative experimentation and mathematical analysis. Molecular Systems Biology, msb e2, D. Forger, M. Drapeau, B. Collins and J. Blau. A new model for circadian clock research? Molecular Systems Biology, msb e1, 2005.

112 References A. Cornish-Bowden and M.L. Cardenas Systems biology may work when we learn to understand the parts in terms of the whole Biochemical Society Transactions, 33, pp , L.H. Hartwell, J.J. Hopfield, S. Leibler and A.W. Murray, From molecular to modular cell biology Nature, 402, pp C47 C50, J. Fisher and T.A. Henzinger, Executable Cell Biology Nature Biotechnology, 25(11), , 2007.