Process algebra and systems biology Vashti Galpin Laboratory for - - PowerPoint PPT Presentation

▶
process algebra and systems biology
SMART_READER_LITE
LIVE PREVIEW

Process algebra and systems biology Vashti Galpin Laboratory for - - PowerPoint PPT Presentation

Sep 23, 2022 239 likes •675 views

Motivation Process algebra PEPA Current research The future Process algebra and systems biology Vashti Galpin Laboratory for Foundations of Computer Science School of Informatics University of Edinburgh 7 December 2007 (Thanks to Jane

slide-1
SLIDE 1

Motivation Process algebra PEPA Current research The future

Process algebra and systems biology

Vashti Galpin Laboratory for Foundations of Computer Science School of Informatics University of Edinburgh 7 December 2007

(Thanks to Jane Hillston and Federica Ciocchetta)

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-2
SLIDE 2

Motivation Process algebra PEPA Current research The future

Motivation

◮ process algebra

◮ different model of computation, reactive system ◮ more explicit model than differential equations ◮ leads to multiple types of analysis Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-3
SLIDE 3

Motivation Process algebra PEPA Current research The future

Motivation

◮ process algebra

◮ different model of computation, reactive system ◮ more explicit model than differential equations ◮ leads to multiple types of analysis

◮ usefulness

◮ for systems biology ◮ for computer science ◮ for this seminar Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-4
SLIDE 4

Motivation Process algebra PEPA Current research The future

Motivation

◮ process algebra

◮ different model of computation, reactive system ◮ more explicit model than differential equations ◮ leads to multiple types of analysis

◮ usefulness

◮ for systems biology ◮ for computer science ◮ for this seminar

◮ survey of existing research

◮ what is a process algebra? ◮ what has been done? ◮ what can be done? Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-5
SLIDE 5

Motivation Process algebra PEPA Current research The future

Process algebra

◮ reactive system

◮ nonterminating, inherently parallel ◮ communicates with environment or other systems ◮ computes by reacting to stimuli Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-6
SLIDE 6

Motivation Process algebra PEPA Current research The future

Process algebra

◮ reactive system

◮ nonterminating, inherently parallel ◮ communicates with environment or other systems ◮ computes by reacting to stimuli

◮ process algebra or process calculus

◮ small number of operators to describe processes, compositional ◮ communication between processes by message passing ◮ mathematical definition of semantics ◮ equivalences equate similar processes ◮ established techniques and tools Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-7
SLIDE 7

Motivation Process algebra PEPA Current research The future

Process algebra

◮ reactive system

◮ nonterminating, inherently parallel ◮ communicates with environment or other systems ◮ computes by reacting to stimuli

◮ process algebra or process calculus

◮ small number of operators to describe processes, compositional ◮ communication between processes by message passing ◮ mathematical definition of semantics ◮ equivalences equate similar processes ◮ established techniques and tools

◮ three main approaches

◮ Communicating Sequential Processes (Hoare, Brooke, Roscoe) ◮ Algebra of Communicating Processes (Baeten, Klop) ◮ Calculus of Communicating Systems (Milner) Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-8
SLIDE 8

Motivation Process algebra PEPA Current research The future

Process algebra (continued)

◮ CSP, denotational semantics

◮ processes mapped to mathematical objects, P ◮ traces, failures, ready sets ◮ equivalence of processes from equality over these objects

P ≡ Q if P = Q

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-9
SLIDE 9

Motivation Process algebra PEPA Current research The future

Process algebra (continued)

◮ CSP, denotational semantics

◮ processes mapped to mathematical objects, P ◮ traces, failures, ready sets ◮ equivalence of processes from equality over these objects

P ≡ Q if P = Q

◮ ACP, algebraic/axiomatic semantics

◮ equations that describe processes with same behaviour

P | Q ≡ Q | P

◮ infer other equivalent processes from equations Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-10
SLIDE 10

Motivation Process algebra PEPA Current research The future

Process algebra (continued)

◮ CCS, operational semantics

◮ rules to describe behaviour of operators ◮ process can perform actions, transitions to other processes ◮ behavioural equivalences defined on labelled transition system ◮ bisimulation Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-11
SLIDE 11

Motivation Process algebra PEPA Current research The future

Process algebra (continued)

◮ CCS, operational semantics

◮ rules to describe behaviour of operators ◮ process can perform actions, transitions to other processes ◮ behavioural equivalences defined on labelled transition system ◮ bisimulation

◮ π-calculus

◮ names, channels and data are not distinguished ◮ can express mobility Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-12
SLIDE 12

Motivation Process algebra PEPA Current research The future

Process algebra (continued)

◮ CCS, operational semantics

◮ rules to describe behaviour of operators ◮ process can perform actions, transitions to other processes ◮ behavioural equivalences defined on labelled transition system ◮ bisimulation

◮ π-calculus

◮ names, channels and data are not distinguished ◮ can express mobility

◮ stochastic process algebra

◮ passing of time associated with transitions, random variable ◮ describes dynamic behaviour and properties ◮ PEPA, Performance Evaluation Process Algebra (Hillston) Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-13
SLIDE 13

Motivation Process algebra PEPA Current research The future

PEPA

◮ CCS-based but uses CSP-type multi-way synchronisation

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-14
SLIDE 14

Motivation Process algebra PEPA Current research The future

PEPA

◮ CCS-based but uses CSP-type multi-way synchronisation ◮ syntax, PEPA model

◮ P ::= (α, r).P

| P + P | P ⊲

⊳

L P

| P/L | C

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-15
SLIDE 15

Motivation Process algebra PEPA Current research The future

PEPA

◮ CCS-based but uses CSP-type multi-way synchronisation ◮ syntax, PEPA model

◮ P ::= (α, r).P

| P + P | P ⊲

⊳

L P

| P/L | C

◮ structured operational semantics, two example rules

(α, r).E

(α,r)

− − − → E E

(α,r)

− − − → E ′ F

(α,r)

− − − → F ′ E ⊲

⊳

L F

(α,r)

− − − → E ′ ⊲

⊳

L F ′

(α ∈ L)

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-16
SLIDE 16

Motivation Process algebra PEPA Current research The future

PEPA

◮ CCS-based but uses CSP-type multi-way synchronisation ◮ syntax, PEPA model

◮ P ::= (α, r).P

| P + P | P ⊲

⊳

L P

| P/L | C

◮ structured operational semantics, two example rules

(α, r).E

(α,r)

− − − → E E

(α,r)

− − − → E ′ F

(α,r)

− − − → F ′ E ⊲

⊳

L F

(α,r)

− − − → E ′ ⊲

⊳

L F ′

(α ∈ L)

◮ can infer transitions using rules, labelled transition system

◮

(α, r).P1 + (β, s).P2 ⊲

⊳

{α} (α, r).Q

(α,r)

− − − → P1 ⊲

⊳

{α} Q

◮

(α, r).P1 + (β, s).P2 ⊲

⊳

{α} (α, r).Q

(β,s)

− − − → P2 ⊲

⊳

{α} (α, r).Q

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-17
SLIDE 17

Motivation Process algebra PEPA Current research The future

Systems biology modelling

◮ general approach (Regev, Silverman, Shapiro)

Concurrency Molecular Metabolism Signal biology transduction Concurrent molecules enzymes and interacting computational processes metabolites proteins Synchronous molecular binding and binding and communication interaction catalysis catalysis

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-18
SLIDE 18

Motivation Process algebra PEPA Current research The future

Systems biology modelling

◮ general approach (Regev, Silverman, Shapiro)

Concurrency Molecular Metabolism Signal biology transduction Concurrent molecules enzymes and interacting computational processes metabolites proteins Synchronous molecular binding and binding and communication interaction catalysis catalysis

◮ molecules as processes or populations as processes?

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-19
SLIDE 19

Motivation Process algebra PEPA Current research The future

Systems biology modelling

◮ general approach (Regev, Silverman, Shapiro)

Concurrency Molecular Metabolism Signal biology transduction Concurrent molecules enzymes and interacting computational processes metabolites proteins Synchronous molecular binding and binding and communication interaction catalysis catalysis

◮ molecules as processes or populations as processes? ◮ stochastic model or deterministic model?

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-20
SLIDE 20

Motivation Process algebra PEPA Current research The future

Systems biology modelling

◮ general approach (Regev, Silverman, Shapiro)

Concurrency Molecular Metabolism Signal biology transduction Concurrent molecules enzymes and interacting computational processes metabolites proteins Synchronous molecular binding and binding and communication interaction catalysis catalysis

◮ molecules as processes or populations as processes? ◮ stochastic model or deterministic model? ◮ aims of modelling

◮ sufficiently faithful ◮ type and tractability of analysis Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-21
SLIDE 21

Motivation Process algebra PEPA Current research The future

Systems biology modelling in PEPA

◮ two modelling approaches

◮ pathway-centric: each subpathway is a process ◮ reagent-centric: each reagent is a process, species focus Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-22
SLIDE 22

Motivation Process algebra PEPA Current research The future

Systems biology modelling in PEPA

◮ two modelling approaches

◮ pathway-centric: each subpathway is a process ◮ reagent-centric: each reagent is a process, species focus

◮ analyses using PEPA model

◮ interpret as continuous time Markov chain (CTMC) ◮ translate to ordinary differential equations (ODEs) ◮ generate a stochastic simulation with Gillespie’s algorithm ◮ model checking of properties using PRISM ◮ find equivalent processes using a behavioural equivalence –

need to find suitable equivalence

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-23
SLIDE 23

Motivation Process algebra PEPA Current research The future

Systems biology modelling in PEPA

◮ two modelling approaches

◮ pathway-centric: each subpathway is a process ◮ reagent-centric: each reagent is a process, species focus

◮ analyses using PEPA model

◮ interpret as continuous time Markov chain (CTMC) ◮ translate to ordinary differential equations (ODEs) ◮ generate a stochastic simulation with Gillespie’s algorithm ◮ model checking of properties using PRISM ◮ find equivalent processes using a behavioural equivalence –

need to find suitable equivalence

◮ modularity, composition, reasoning

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-24
SLIDE 24

Motivation Process algebra PEPA Current research The future

Systems biology modelling in PEPA

◮ two modelling approaches

◮ pathway-centric: each subpathway is a process ◮ reagent-centric: each reagent is a process, species focus

◮ analyses using PEPA model

◮ interpret as continuous time Markov chain (CTMC) ◮ translate to ordinary differential equations (ODEs) ◮ generate a stochastic simulation with Gillespie’s algorithm ◮ model checking of properties using PRISM ◮ find equivalent processes using a behavioural equivalence –

need to find suitable equivalence

◮ modularity, composition, reasoning ◮ refinement, abstraction, causality

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-25
SLIDE 25

Motivation Process algebra PEPA Current research The future

Example

◮ example, single substrate enzyme catalyzed reaction

E + S

bind

− − − − ⇀ ↽ − − − −

unbind ES produce

− − − − → E + P

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-26
SLIDE 26

Motivation Process algebra PEPA Current research The future

Example

◮ example, single substrate enzyme catalyzed reaction

E + S

bind

− − − − ⇀ ↽ − − − −

unbind ES produce

− − − − → E + P

◮ reagent-centric PEPA model

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-27
SLIDE 27

Motivation Process algebra PEPA Current research The future

Example

◮ example, single substrate enzyme catalyzed reaction

E + S

bind

− − − − ⇀ ↽ − − − −

unbind ES produce

− − − − → E + P

◮ reagent-centric PEPA model ◮ high and low concentrations, discretized

Sh

def

= (b, rb).Sℓ Sℓ

def

= (u, ru).Sh Eh

def

= (b, rb).Eℓ Eℓ

def

= (u, ru).Eh + (p, rp).Eh ESℓ

def

= (b, rb).ESh ESh

def

= (u, ru).ESℓ + (p, rp).ESℓ Pℓ

def

= (p, rp).Ph

  • Sh ⊲

⊳

{b,u} Eh

⊲

⊳

{b,u,p} ESℓ

  • ⊲

⊳

{p} Pℓ Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-28
SLIDE 28

Motivation Process algebra PEPA Current research The future

PEPA

◮ biochemical signalling pathways

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-29
SLIDE 29

Motivation Process algebra PEPA Current research The future

PEPA

◮ biochemical signalling pathways ◮ model of the influence of RKIP on the ERK signalling

pathway (Calder, Gillmore and Hillston)

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-30
SLIDE 30

Motivation Process algebra PEPA Current research The future

PEPA

◮ biochemical signalling pathways ◮ model of the influence of RKIP on the ERK signalling

pathway (Calder, Gillmore and Hillston)

◮ model of MAP Kinase cascade (Calder, Duguid, Gilmore,

Hillston)

◮ original Schoeberl model based on ODEs, Matlab analysis ◮ PEPA model created, extracted ODEs, matched results ◮ stochastic simulation results differed from ODE results ◮ step size for ODE analysis too large, peak missed Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-31
SLIDE 31

Motivation Process algebra PEPA Current research The future

PEPA

◮ biochemical signalling pathways ◮ model of the influence of RKIP on the ERK signalling

pathway (Calder, Gillmore and Hillston)

◮ model of MAP Kinase cascade (Calder, Duguid, Gilmore,

Hillston)

◮ original Schoeberl model based on ODEs, Matlab analysis ◮ PEPA model created, extracted ODEs, matched results ◮ stochastic simulation results differed from ODE results ◮ step size for ODE analysis too large, peak missed

◮ new process algebra, Bio-PEPA (Ciocchetta, Hillston)

◮ reagent-centric, includes stoichiometry ◮ better modelling of reaction rates, general kinetic laws ◮ new syntax, parameterised Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-32
SLIDE 32

Motivation Process algebra PEPA Current research The future

Other process algebra approaches

◮ stochastic π-calculus (Regev, Shapiro, Silverman, Priami,

Cardelli)

◮ binary communication only, individual-based, mobility ◮ analysis by stochastic simulation ◮ models of metabolic pathways, gene transcription, signal

transduction

◮ cell cycle control in eukaryotes (Lecca, Priami) ◮ lymphocyte-endothelial interactions in inflamed brain venules

(Lecca, Priami, Laudanna, Constantin)

◮ VIrtual CEll (VICE) (Chiarugi, Curti, Degano, Marangoni) Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-33
SLIDE 33

Motivation Process algebra PEPA Current research The future

Other process algebra approaches

◮ stochastic π-calculus (Regev, Shapiro, Silverman, Priami,

Cardelli)

◮ binary communication only, individual-based, mobility ◮ analysis by stochastic simulation ◮ models of metabolic pathways, gene transcription, signal

transduction

◮ cell cycle control in eukaryotes (Lecca, Priami) ◮ lymphocyte-endothelial interactions in inflamed brain venules

(Lecca, Priami, Laudanna, Constantin)

◮ VIrtual CEll (VICE) (Chiarugi, Curti, Degano, Marangoni)

◮ Brane Calculus (Cardelli, Danos, Pradalier)

◮ deals with spatial aspects ◮ models membranes explicitly ◮ description of virus infection Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-34
SLIDE 34

Motivation Process algebra PEPA Current research The future

Other process algebra approaches (continued)

◮ beta binders (Priami, Quaglia)

◮ beta boxes contain π-calculus terms ◮ beta boxes have external sites for interaction with others ◮ very detailed ◮ affinity between sites can be defined ◮ potential use in drug discovery Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-35
SLIDE 35

Motivation Process algebra PEPA Current research The future

Other process algebra approaches (continued)

◮ beta binders (Priami, Quaglia)

◮ beta boxes contain π-calculus terms ◮ beta boxes have external sites for interaction with others ◮ very detailed ◮ affinity between sites can be defined ◮ potential use in drug discovery

◮ Bio-Ambients (Regev, Panina, Silverman, Cardelli, Shapiro)

◮ based on ambient calculus, extension of π-calculus ◮ movement, location, compartments ◮ hypothalamic weight regulation system, multiple levels ◮ simulation results give support for model Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-36
SLIDE 36

Motivation Process algebra PEPA Current research The future

Other process algebra approaches (continued)

◮ beta binders (Priami, Quaglia)

◮ beta boxes contain π-calculus terms ◮ beta boxes have external sites for interaction with others ◮ very detailed ◮ affinity between sites can be defined ◮ potential use in drug discovery

◮ Bio-Ambients (Regev, Panina, Silverman, Cardelli, Shapiro)

◮ based on ambient calculus, extension of π-calculus ◮ movement, location, compartments ◮ hypothalamic weight regulation system, multiple levels ◮ simulation results give support for model

◮ κ-calculus (Danos, Laneve)

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-37
SLIDE 37

Motivation Process algebra PEPA Current research The future

The future

◮ process algebra as modelling technique for systems biology

◮ build formal models with explicit interaction and

compositionality using simple but descriptive language

◮ many types of analysis from one syntactic description ◮ provide insights, generate hypotheses ◮ allow experimentation in silico ◮ multi-scale (concentrations vary), stiff (reaction rates vary) ◮ compositionality to model different levels ◮ techniques for insufficient data, abstraction ◮ provision of efficient software tools Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-38
SLIDE 38

Motivation Process algebra PEPA Current research The future

The future

◮ process algebra as modelling technique for systems biology

◮ build formal models with explicit interaction and

compositionality using simple but descriptive language

◮ many types of analysis from one syntactic description ◮ provide insights, generate hypotheses ◮ allow experimentation in silico ◮ multi-scale (concentrations vary), stiff (reaction rates vary) ◮ compositionality to model different levels ◮ techniques for insufficient data, abstraction ◮ provision of efficient software tools

◮ gold standard: biological insights achieved through use of

process algebra that are not possible through existing approaches

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-39
SLIDE 39

Motivation Process algebra PEPA Current research The future

The future (continued)

◮ process algebra as model/metaphor for systems biology

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-40
SLIDE 40

Motivation Process algebra PEPA Current research The future

The future (continued)

◮ process algebra as model/metaphor for systems biology ◮ Noble’s ten principles of systems biology

  • 1. Biological functionality is multi-level
  • 2. Transmission of information is not one way
  • 3. DNA is not the sole transmitter of inheritance
  • 4. There is no privileged level of causality
  • 5. Gene ontology will fail without higher-level insight
  • 6. There is no genetic program
  • 7. There are no programs at any other level
  • 8. There are no programs in the brain
  • 9. The self is not an object
  • 10. There are many more to be discovered

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-41
SLIDE 41

Motivation Process algebra PEPA Current research The future

The future (continued)

◮ process algebra as model/metaphor for systems biology ◮ Noble’s ten principles of systems biology

  • 1. Biological functionality is multi-level
  • 2. Transmission of information is not one way
  • 3. DNA is not the sole transmitter of inheritance
  • 4. There is no privileged level of causality
  • 5. Gene ontology will fail without higher-level insight
  • 6. There is no genetic program
  • 7. There are no programs at any other level
  • 8. There are no programs in the brain
  • 9. The self is not an object
  • 10. There are many more to be discovered

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology

slide-42
SLIDE 42

Motivation Process algebra PEPA Current research The future

The future (continued)

◮ process algebra as model/metaphor for systems biology ◮ Noble’s ten principles of systems biology

  • 1. Biological functionality is multi-level
  • 2. Transmission of information is not one way
  • 3. DNA is not the sole transmitter of inheritance
  • 4. There is no privileged level of causality
  • 5. Gene ontology will fail without higher-level insight
  • 6. There is no genetic program
  • 7. There are no programs at any other level
  • 8. There are no programs in the brain
  • 9. The self is not an object
  • 10. There are many more to be discovered

◮ reactive system computing is a more general model than

program-as-a-function computing

Vashti Galpin, LFCS, University of Edinburgh Process algebra and systems biology