Computational Methods for Systems and Synthetic Biology Franois - - PowerPoint PPT Presentation

▶
computational methods for
SMART_READER_LITE
LIVE PREVIEW

Computational Methods for Systems and Synthetic Biology Franois - - PowerPoint PPT Presentation

Oct 10, 2023 859 likes •1.67k views

Computational Methods for Systems and Synthetic Biology Franois Fages The French National Institute for Research in Computer Science and Control INRIA Paris-Rocquencourt Constraint Programming Group http://contraintes.inria.fr/ Franois

slide-1
SLIDE 1

24/06/2013 François Fages – Ecole Jeunes Chercheurs - Porquerolles 1

Computational Methods for Systems and Synthetic Biology

François Fages

The French National Institute for Research in Computer Science and Control INRIA Paris-Rocquencourt

Constraint Programming Group

http://contraintes.inria.fr/

slide-2
SLIDE 2

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 2

Overview of the Lectures

1. Introduction

  • Transposing concepts from programming to the analysis of living processes

2. Rule-based Modeling in Biocham

  • Macromolecules, compartments and elementary processes in the cell
  • Boolean, Differential and Stochastic interpretations of reaction rules
  • Cell signaling, Gene expression, Retrovirus, Cell cycle

3. Temporal Logic constraints in Biocham

  • Qualitative properties in propositional Computation Tree Logic CTL
  • Quantitative properties in quantifier-free Linear Time Logic LTL(R)
  • Parameter optimization and robustness w.r.t. temporal logic properties

4. Conclusion 5. Killer lecture: abstract interpretation in Biocham

slide-3
SLIDE 3

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 3

References

A wonderful textbook: Molecular Cell Biology. 5th Edition, 1100 pages+CD, Freeman Publ. Lodish, Berk, Zipursky, Matsudaira, Baltimore, Darnell. Nov. 2003. Formal Cell Biology in BIOCHAM (tutorial). François Fages and Sylvain Soliman. 8th International School on Computational Systems Biology. ISpringer-Verlag, LNCS 5016. Mar. 2008.(pdf) The Biochemical Abstract Machine BIOCHAM. http://contraintes.inria.fr/BIOCHAM Modeling dynamic phenomena in molecular and cellular biology.

  • Segel. Cambridge Univ. Press. 1987.
slide-4
SLIDE 4

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 4

Systems Biology ?

“Systems Biology aims at systems-level understanding [which] requires a set of principles and methodologies that links the behaviors of molecules to systems characteristics and functions.”

  • H. Kitano, ICSB 2000
  • Analyze (post-)genomic data produced with high-throughput technologies

(stored in databases like GO, KEGG, BioCyc, etc.);

  • Integrate heterogeneous data about a specific problem;
  • Understand and predict the behaviors of large networks of genes and

proteins.  Systems Biology Markup Language (SBML): model exchange format  SBML model repositories: e.g. biomodels.net 261 curated models

slide-5
SLIDE 5

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 5

Issue of Abstraction in Systems Biology

Models are built in Systems Biology with two contradictory perspectives :

slide-6
SLIDE 6

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 6

Issue of Abstraction in Systems Biology

Models are built in Systems Biology with two contradictory perspectives : 1) Models for representing knowledge : the more concrete the better

slide-7
SLIDE 7

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 7

Issue of Abstraction in Systems Biology

Models are built in Systems Biology with two contradictory perspectives : 1) Models for representing knowledge : the more concrete the better 2) Models for making predictions : the more abstract the better !

slide-8
SLIDE 8

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 8

Issue of Abstraction in Systems Biology

Models are built in Systems Biology with two contradictory perspectives : 1) Models for representing knowledge : the more concrete the better 2) Models for making predictions : the more abstract the better !

La simplicité est la sophistication suprême. Léonard de Vinci.

slide-9
SLIDE 9

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 9

Issue of Abstraction in Systems Biology

Models are built in Systems Biology with two contradictory perspectives : 1) Models for representing knowledge : the more concrete the better 2) Models for making predictions : the more abstract the better !

La simplicité est la sophistication suprême. Léonard de Vinci.

These perspectives can be reconciled by organizing formalisms and models into hierarchies of abstractions. To understand a system is not to know everything about it but to know abstraction levels that are sufficient for answering questions about it

slide-10
SLIDE 10

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 10

Issue of Abstraction in Systems Biology

Models are built in Systems Biology with two contradictory perspectives : 1) Models for representing knowledge : the more concrete the better 2) Models for making predictions : the more abstract the better !

La simplicité est la sophistication suprême. Léonard de Vinci.

These perspectives can be reconciled by organizing formalisms and models into hierarchies of abstractions. To understand a system is not to know everything about it but to know abstraction levels that are sufficient for answering questions about it

Karl Popper empirical falsification versus positivist confirmation

slide-11
SLIDE 11

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 11

Formal Semantics of Living Processes ?

Formally, “the” behavior of a system depends on our choice of observables. ? ?

Mitosis movie [Lodish et al. 03]

slide-12
SLIDE 12

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 12

Boolean Semantics

  • Formally, “the” behavior of a system depends on our choice of observables.
  • Presence/absence of molecules
  • Boolean transitions models

1

Mitosis movie [Lodish et al. 03]

slide-13
SLIDE 13

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 13

Continuous Semantics

  • Formally, “the” behavior of a system depends on our choice of observables.
  • Concentrations of molecules
  • Ordinary Differential Equation models

x ý

Mitosis movie [Lodish et al. 03]

slide-14
SLIDE 14

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 14

Stochastic Semantics

  • Formally, “the” behavior of a system depends on our choice of observables.
  • Numbers of molecules
  • Continuous Time Markov Chain models

n 

Mitosis movie [Lodish et al. 03]

slide-15
SLIDE 15

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 15

Temporal Logic LTL

  • Formally, “the” behavior of a system depends on our choice of observables.
  • Presence/absence of molecules
  • Temporal logic formulas

F x F x F (x ^ F ( x ^ y)) FG (x v y) …

Mitosis movie [Lodish et al. 03]

slide-16
SLIDE 16

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 16

Temporal Logic LTL(R)

  • Formally, “the” behavior of a system depends on our choice of observables.
  • Concentrations of molecules
  • Temporal Logic with Constraints over R

F x>1 F (x >0.2) F (x >0.2 ^ F (x<0.1 ^ y>0.2)) FG (x>0.2 v y>0.2) …

Mitosis movie [Lodish et al. 03]

slide-17
SLIDE 17

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 17

Hierarchy of Semantics

Stochastic model Differential model Discrete model

abstraction concretization

Boolean model

Theory of abstract Interpretation Abstractions as Galois connections [Cousot Cousot POPL’77] [Fages Soliman CMSB’06,TCS’07]

Syntactical model

slide-18
SLIDE 18

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 18

Regulation Graph as Abstraction

Stochastic model Differential model Discrete model

abstraction concretization

Boolean model Syntactical model

[Fages Soliman FMSB’06]

Structural regulation graph (pos/neg influences w.r.t. the stoichiometric coefficient in reactions)

  • Thm. Same graphs for

monotonic kinetics Differential regulation graph (pos/neg influences w.r.t. the signs of the Jacobian)

slide-19
SLIDE 19

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 19

Regulation Graphs Circuit Analyses

Stochastic model Differential model Discrete model

abstraction concretization

Boolean model Syntactical model Jacobian circuit analysis Discrete circuit analysis Boolean circuit analysis

abstraction abstraction abstraction

  • Thm. Positive circuits are

a necessary condition for multistationarity

[Thomas 81] [Soulé 03] [Remy Ruet Thieffry 05] [Richard 07] [Soliman 13]

slide-20
SLIDE 20

24/06/2013 François Fages - Ecoles Jeunes Chercheurs - Porquerolles 20

Mammalian Cell Cycle Control Map [Kohn 99]

slide-21
SLIDE 21

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 21

Hierachy of Models / Model Reductions

Models of circadian clock in http://www.biomodels.net Reductions as Subgraph Epimorphisms [Gay Fages Soliman ECCB’10 DAM 13]

slide-22
SLIDE 22

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 22

Overview of the Lectures

1. Introduction

  • Transposing concepts from programming to the analysis of living processes

2. Rule-based Modeling in Biocham

  • Macromolecules, compartments and elementary processes in the cell
  • Boolean, Differential and Stochastic interpretations of reaction rules
  • Cell signaling, cell cycle models

3. Temporal Logic constraints in Biocham

  • Qualitative properties in propositional Computation Tree Logic CTL
  • Quantitative properties in quantifier-free Linear Time Logic LTL(R)
  • Model inference from temporal logic properties

4. Conclusion 5. Killer lecture: abstract interpretation in Biocham

slide-23
SLIDE 23

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 23

Cell Molecules

  • Small molecules: covalent bonds 50-200 kcal/mol

– 70% water – 1% ions – 6% amino acids (20), nucleotides (5), – fats, sugars, ATP, ADP, …

slide-24
SLIDE 24

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 24

Cell Molecules

  • Small molecules: covalent bonds 50-200 kcal/mol

– 70% water – 1% ions – 6% amino acids (20), nucleotides (5), – fats, sugars, ATP, ADP, …

  • Macromolecules: hydrogen bonds, ionic, hydrophobic, Waals 1-5 kcal/mol

Stability and bindings determined by the number of weak bonds: 3D shape – 20% proteins (50-104 amino acids) – DNA (102-106 nucleotides AGCT) – RNA (102-104 nucleotides AGCU)

slide-25
SLIDE 25

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 25

DNA Deoxyribonucleic Acid

1) Primary structure: word over 4 nucleotides Adenine, Guanine, Cytosine, Thymine

slide-26
SLIDE 26

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 26

DNA Deoxyribonucleic Acid

1) Primary structure: word over 4 nucleotides Adenine, Guanine, Cytosine, Thymine 2) Secondary structure: double helix of pairs A--T and C---G stabilized by hydrogen bonds Size of DNA = number of pairs A gene is a subword of DNA Nobel Prizes Watson and Crick (1956)

slide-27
SLIDE 27

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 27

DNA: Genome Size

Species Genome size Chromosomes Coding DNA

  • E. Coli (bacteria)

5 Mb 1 circular 100 % 12 Mb … 3 Gb … 15 Gb … 140 Gb Artificial life by Craig Venter: fully synthetic bacteria genome (0,39 $/b) implemented in a bacteria without DNA still living and proliferating!

slide-28
SLIDE 28

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 28

DNA: Genome Size

Species Genome size Chromosomes Coding DNA

  • E. Coli (bacteria)

5 Mb 1 circular 100 %

  • S. Cerevisae (yeast)

12 Mb 16 70 % … 3 Gb … 15 Gb … 140 Gb

slide-29
SLIDE 29

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 29

DNA: Genome Size

Species Genome size Chromosomes Coding DNA

  • E. Coli (bacteria)

5 Mb 1 circular 100 %

  • S. Cerevisae (yeast)

12 Mb 16 70 % Mouse, Human 3 Gb 20, 23 15 % … 15 Gb … 140 Gb

3,200,000,000 pairs of nucleotides Fully sequenced in year 2000 Single nucleotide polymorphism =1/2kb

slide-30
SLIDE 30

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 30

Genome Size

Species Genome size Chromosomes Coding DNA

  • E. Coli (bacteria)

4 Mb 1 100 %

  • S. Cerevisae (yeast)

12 Mb 16 70 % Mouse, Human 3 Gb 20, 23 15 % Onion 15 Gb 8 1 % … 140 Gb

slide-31
SLIDE 31

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 31

Genome Size

Species Genome size Chromosomes Coding DNA

  • E. Coli (bacteria)

4 Mb 1 100 %

  • S. Cerevisae (yeast)

12 Mb 16 70 % Mouse, Human 3 Gb 20, 23 15 % Onion 15 Gb 8 1 % Lungfish 140 Gb 0.7 %

slide-32
SLIDE 32

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 32

DNA Replication

  • 1. Separation of the two helices
  • 2. Production of one complementary strand for each copy

(from one or several starting points of replication)

  • 3. Segregation
  • 4. Mitosis (cell division)
slide-33
SLIDE 33

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 33

Gene Transcription and Translation

  • 1. Activation (Inhibition): Nobel prize Jakob and Monod (1965)

transcription factors bind to the regulatory region of the gene

  • 2. Transcription:

RNA polymerase copies the DNA from start to stop positions into a single stranded pre-mature messenger pRNA

  • 3. (Alternative) splicing:

non coding regions of pRNA are removed giving mature messenger mRNA

  • 4. Translation:

mRNA moves to cytoplasm and binds to ribosome to assemble a protein

slide-34
SLIDE 34

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 34

Formal Genes: Syntax

  • Part of DNA, unique E2
  • Activation

binding of promotion factor E2-(E2f13-DP12)

  • Repression (inhibition)

binding of another molecule E2-Rep

slide-35
SLIDE 35

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 35

Transcription and Translation Rules

Activation E2 + E2f13-DP12 => E2-E2f13-DP12 Repression E2 + Rep => E2-Rep

slide-36
SLIDE 36

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 36

Transcription and Translation Rules

Activation E2 + E2f13-DP12 => E2-E2f13-DP12 Repression E2 + Rep => E2-Rep Transcription _ =[E2-E2F13-DP12]=> pRNAcycA

slide-37
SLIDE 37

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 37

Transcription and Translation Rules

Activation E2 + E2f13-DP12 => E2-E2f13-DP12 Repression E2 + Rep => E2-Rep Transcription _ =[E2-E2F13-DP12]=> pRNAcycA (Alternative) Splicing pRNAcycA => mRNAcycA (pRNAcycA => mRNAcycA2)

slide-38
SLIDE 38

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 38

Transcription and Translation Rules

Activation E2 + E2f13-DP12 => E2-E2f13-DP12 Repression E2 + Rep => E2-Rep Transcription _ =[E2-E2F13-DP12]=> pRNAcycA (Alternative) Splicing pRNAcycA => mRNAcycA (pRNAcycA => mRNAcycA2) Translation mRNAcycA => mRNAcycA::cyt mRNAcycA::cyt + ribosome::cyt => cycA::cyt + ribosome::cyt

(mRNAcycA2::cyt + ribosome::cyt => cycA2::cyt + ribosome::cyt)

slide-39
SLIDE 39

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 39

Peptides and Proteins

1) Primary structure: word of n amino acids residues (20n possibilities) linked with C-N bonds

slide-40
SLIDE 40

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 40

Peptides and Proteins

1) Primary structure: word of n amino acids residues (20n possibilities) linked with C-N bonds Example: MPRI Methionine-Proline-Arginine-Isoleucine

slide-41
SLIDE 41

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 41

Peptides and Proteins

1) Primary structure: word of n amino acids residues (20n possibilities) linked with C-N bonds Example: MPRI Methionine-Proline-Arginine-Isoleucine Question: How many nucleotides on the gene are necessary to code for one amino acid on the protein?

slide-42
SLIDE 42

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 42

Peptides and Proteins

1) Primary structure: word of n amino acids residues (20n possibilities) linked with C-N bonds Example: MPRI Methionine-Proline-Arginine-Isoleucine Question: How many nucleotides on the gene are necessary to code for one amino acid on the protein? 3 ! 2 are not sufficient: 4*4=16 3 is sufficient but with a lot of non coding sequences: 4*4*4=64

slide-43
SLIDE 43

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 43

Peptides and Proteins

1) Primary structure: word of n amino acids residues (20n possibilities) linked with C-N bonds Example: MPRI Methionine-Proline-Arginine-Isoleucine 2) Secondary: word of m a-helix, b-strands, random coils,… (3m-10m) stabilized by hydrogen bonds H---O

slide-44
SLIDE 44

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 44

Peptides and Proteins

1) Primary structure: word of n amino acids residues (20n possibilities) linked with C-N bonds Example: MPRI Methionine-Proline-Arginine-Isoleucine 2) Secondary: word of m a-helix, b-strands, random coils,… (3m-10m) stabilized by hydrogen bonds H---O 3) Tertiary 3D structure: spatial folding stabilized by hydrophobic interactions explains the protein interaction capabilities

slide-45
SLIDE 45

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 45

Formal Proteins: Syntax

  • Cyclin dependent kinase 1 Cdk1

(free, inactive)

slide-46
SLIDE 46

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 46

Formal Proteins: Syntax

  • Cyclin dependent kinase 1 Cdk1

(free, inactive)

  • Complex Cdk1-Cyclin B Cdk1–CycB

(low activity)

slide-47
SLIDE 47

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 47

Formal Proteins: Syntax

  • Cyclin dependent kinase 1 Cdk1

(free, inactive)

  • Complex Cdk1-Cyclin B Cdk1–CycB

(low activity)

  • Phosphorylated form Cdk1~{thr161}-CycB

at site threonine 161 (high activity)

slide-48
SLIDE 48

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 48

Formal Proteins

  • Cyclin dependent kinase 1 Cdk1

(free, inactive)

  • Complex Cdk1-Cyclin B Cdk1–CycB

(low activity)

  • Phosphorylated form Cdk1~{thr161}-CycB

at site threonine 161 (high activity) “Mitosis-Promoting Factor” phosphorylates actin in microtubules  nuclear division

slide-49
SLIDE 49

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 49

Seven Main Reaction Schemas

  • Complexation: A + B => A-B. Decomplexation A-B => A + B.

cdk1+cycB => cdk1–cycB

  • Phosphorylation: A =[C]=> A~{p}. Dephosphorylation A~{p} =[C]=> A.

Cdk1-CycB =[Myt1]=> Cdk1~{thr161}-CycB Cdk1~{thr14,tyr15}-CycB =[Cdc25~{Nterm}]=> Cdk1-CycB

  • Synthesis: _ =[C]=> A. Degradation: A =[C]=> _.

_=[E2-E2f13-Dp12]=>cycA cycE =[@UbiPro]=> _ (not for cycE-cdk2 which is stable)

  • Transport: A::L1 => A::L2.

Cdk1~{p}-CycB::cytoplasm=>Cdk1~{p}-CycB::nucleus

slide-50
SLIDE 50

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 50

Syntax of Objects

E == entity | E-E | E~{p1,…,pn}

  • Entity: molecule name, gene binding site,…
  • - : binding operator for protein complexes, gene binding sites, …

associative and commutative.

  • ~{…}: set of modified sites (phosphorylation, acetylation, methylation,…)

O == E | E::location

  • Object: entity localized in a symbolic compartment (nucleus, cytoplasm,

membrane, …)

slide-51
SLIDE 51

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 51

Syntax of Reactions

S ::= _ | k*O + S A solution is a of objects 0N, written as a linear expression R ::= S => S | f for R A reaction is a multiset rewriting rule, possibly given with a kinetic expression f A =[C]=> B stands for A+C => B+C A <=> B stands for A=>B and B=>A, Syntax compatible with the XML syntax of Systems Biology Markup Language SBML, widely used exchange format for reaction models.

slide-52
SLIDE 52

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 52

From Syntax to Semantics

  • 1. Boolean Semantics: presence-absence of molecules

– Boolean Transition System (asynchronous, synchronous)

  • 2. Discrete semantics: numbers of molecules

– Petri net

  • 3. Continuous Semantics: concentrations

– Ordinary Differential Equations – Hybrid automata

  • 4. Stochastic Semantics: numbers of molecules

– Continuous time Markov chain

slide-53
SLIDE 53

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 53

Cell Signaling

  • Signals:

– hormones: insulin, adrenaline, steroids, EGF, …, – neighboring cell membrane proteins: Delta – nutriments, light, pressure, …

slide-54
SLIDE 54

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 54

Cell Signaling

  • Signals:

– hormones: insulin, adrenaline, steroids, EGF, …, – neighboring cell membrane proteins: Delta – nutriments, light, pressure, …

  • Receptors transmembrane proteins:

– Tyrosine kinases, – G protein-coupled, – TGFβ, – Notch, – Ionic channels – …

slide-55
SLIDE 55

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 55

Receptor Tyrosine Kinase RTK

L + R <=> L-R L-R + L-R => L-R-L-R RAS-GDP =[L-R-L-R]=> RAS-GTP

slide-56
SLIDE 56

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 56

MAPK Signaling Pathways

  • Input:

RAS activated by the receptor activates RAF RAS-GTP + RAF-P14-3-3 => RAS-GDP + RAF + P14-3-3

  • Output:

active MAPK moves to the nucleus phosphorylates a transcription factor which stimulates gene expression RAF + … => … … => MAPK~{T183,Y185}

slide-57
SLIDE 57

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 57

Five MAP Kinase Pathways in Budding Yeast (Saccharomyces Cerevisiae)

slide-58
SLIDE 58

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 58

Three Levels MAPK Cascade

slide-59
SLIDE 59

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 59

Reaction Model of the MAPK Cascade

[Levchenko et al. PNAS 2000]

(MA(1), MA(0.4)) for RAF + RAFK <=> RAF-RAFK. (MA(0.5),MA(0.5)) for RAF~{p1} + RAFPH <=> RAF~{p1}-RAFPH. (MA(3.3),MA(0.42)) for MEK~$P + RAF~{p1} <=> MEK~$P-RAF~{p1} where p2 not in $P. (MA(10),MA(0.8)) for MEKPH + MEK~{p1}~$P <=> MEK~{p1}~$P-MEKPH. (MA(20),MA(0.7)) for MAPK~$P + MEK~{p1,p2} <=> MAPK~$P-MEK~{p1,p2} where p2 not in $P. (MA(5),MA(0.4)) for MAPKPH + MAPK~{p1}~$P <=> MAPK~{p1}~$P-MAPKPH. MA(0.1) for RAF-RAFK => RAFK + RAF~{p1}. MA(0.1) for RAF~{p1}-RAFPH => RAF + RAFPH. MA(0.1) for MEK~{p1}-RAF~{p1} => MEK~{p1,p2} + RAF~{p1}. MA(0.1) for MEK-RAF~{p1} => MEK~{p1} + RAF~{p1}. MA(0.1) for MEK~{p1}-MEKPH => MEK + MEKPH. MA(0.1) for MEK~{p1,p2}-MEKPH => MEK~{p1} + MEKPH. MA(0.1) for MAPK-MEK~{p1,p2} => MAPK~{p1} + MEK~{p1,p2}. MA(0.1) for MAPK~{p1}-MEK~{p1,p2} => MAPK~{p1,p2} + MEK~{p1,p2}. MA(0.1) for MAPK~{p1}-MAPKPH => MAPK + MAPKPH. MA(0.1) for MAPK~{p1,p2}-MAPKPH => MAPK~{p1} + MAPKPH.

slide-60
SLIDE 60

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 60

Reaction Hypergraph

Bipartite Proteins-Reactions Graph

GraphViz

http://www.research.att.co/sw/tools/graphviz

slide-61
SLIDE 61

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 61

Influence Graph inferred from the reaction graph of the MAPK “cascade” Negative circuit [Fages Soliman CMSB 06] Possibility of oscillations [Qiao et al. PLOS 07]

Influence Graph

slide-62
SLIDE 62

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 62

Differential Simulation

slide-63
SLIDE 63

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 63

Stochastic Simulation

slide-64
SLIDE 64

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 64

Boolean Simulation

slide-65
SLIDE 65

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 65

Semantics of Reaction Models

Reaction rule: k*[A]*[B] for A+B => C Three interpretations in Biocham:

  • 1. Stochastic Semantics: number of molecules

– Continuous time Markov chain

  • 2. Differential Semantics: concentration

– Ordinary Differential Equations – Hybrid automata

  • 3. Boolean Semantics: presence-absence of molecules

– Asynchronous Transition System A & B  C & A/A & B/B

slide-66
SLIDE 66

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 66

Automatic Generation of CTL Properties

reachable(MAPK~{p1})) reachable(!(MAPK~{p1})))

  • scil(MAPK~{p1}))

… reachable(MAPKPH-MAPK~{p1})) reachable(!(MAPKPH-MAPK~{p1})))

  • scil(MAPKPH-MAPK~{p1}))

AG(!(MAPKPH-MAPK~{p1})->checkpoint(MAPKPH,MAPKPH-MAPK~{p1}))) AG(!(MAPKPH-MAPK~{p1})->checkpoint(MAPK~{p1},MAPKPH-MAPK~{p1}))) … reachable(MAPK~{p1,p2})) reachable(!(MAPK~{p1,p2})))

  • scil(MAPK~{p1,p2}))

…

slide-67
SLIDE 67

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 67

Model Reductions Preserving CTL Properties

After reduction, 20 rules remain. Deletions: RAF-RAFK=>RAF+RAFK RAFPH-RAF~{p1}=>RAFPH+RAF~{p1} MEK-RAF~{p1}=>MEK+RAF~{p1} MEKPH-MEK~{p1}=>MEKPH+MEK~{p1} MAPK-MEK~{p1,p2}=>MAPK+MEK~{p1,p2} MAPKPH-MAPK~{p1}=>MAPKPH+MAPK~{p1} MEK~{p1}-RAF~{p1}=>MEK~{p1}+RAF~{p1} MEKPH-MEK~{p1,p2}=>MEKPH+MEK~{p1,p2} MAPK~{p1}-MEK~{p1,p2}=>MAPK~{p1}+MEK~{p1,p2} MAPKPH-MAPK~{p1,p2}=>MAPKPH+MAPK~{p1,p2}

slide-68
SLIDE 68

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 68

Model Reduction as Subgraph Epimorphism

011_levc

MAPK models from the SBML model repository http://www.biomodels.net Reaction (hyper)graph reduction G1G2 through 4 operations: molecule/reaction merge/deletion iff SEPI from G1 to G2

[Gay Fages Soliman 2010 ECCB, Bioinformatics, DAM 2013]

slide-69
SLIDE 69

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 69

Biological process model = Concurrent Transition System Biological property = Temporal Logic Formula Biological validation = Model-checking Model inference = Temporal Constraint Solving

  • [Lincoln et al. PSB’02] [Chabrier Fages CMSB’03] [Bernot et al. TCS’04] …

Model: BIOCHAM Biological Properties:

  • Boolean - simulation - Temporal logic CTL
  • Differential - query evaluation - LTL(R), QFLTL(R) constraints
  • Stochastic - rule learning - CSL

(SBML) - parameter search

  • Influences

A Logical Paradigm for Systems Biology

slide-70
SLIDE 70

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 70

A Programmer View at Cell Computations

Size of genome

  • 5 Mb for bacteria: normal size program (Biocham binary: 15Mb as yeast)
slide-71
SLIDE 71

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 71

A Programmer View at Cell Computations

Size of genome

  • 5 Mb for bacteria: normal size program (Biocham binary: 15Mb as yeast)
  • 3 Gb for human: normal size of a video not for a program
slide-72
SLIDE 72

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 72

A Programmer View at Cell Computations

Size of genome

  • 5 Mb for bacteria: normal size program (Biocham binary: 15Mb as yeast)
  • 3 Gb for human: normal size of a video not for a program
  • 140 Gb for lung fish: nature error !
slide-73
SLIDE 73

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 73

A Programmer View at Cell Computations

Size of genome

  • 5 Mb for bacteria: normal size program (Biocham binary: 15Mb as yeast)
  • 3 Gb for human: normal size of a video not for a program
  • 140 Gb for lung fish: nature error !

Speed of interactions

  • Protein interactions: enzyme-substrate collisions at 0,5 Mhz, quite slow
  • Gene expression: hours ! as long as reinstalling an operating system
slide-74
SLIDE 74

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 74

A Programmer View at Cell Computations

Size of genome

  • 5 Mb for bacteria: normal size program (Biocham binary: 15Mb as yeast)
  • 3 Gb for human: normal size of a video not for a program
  • 140 Gb for lung fish: nature error !

Speed of interactions

  • Protein interactions: enzyme-substrate collisions at 0,5 Mhz, quite slow
  • Gene expression: hours ! as long as reinstalling an operating system

Concurrent computation paradigm

  • Chemical metaphor for concurrent programming [Banatre, Le Metayer 78]
  • CHAM [Berry 85] to express the operational semantics of the Pi-Calculus
  • Membranes for modules: just like cell compartments
slide-75
SLIDE 75

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 75

A Programmer View at Cell Computations

Size of genome

  • 5 Mb for bacteria: normal size program (Biocham binary: 15Mb as yeast)
  • 3 Gb for human: normal size of a video not for a program
  • 140 Gb for lung fish: nature error !

Speed of interactions

  • Protein interactions: enzyme-substrate collisions at 0,5 Mhz, quite slow
  • Gene expression: hours ! as long reinstalling an operating system

Concurrent computation paradigm

  • Chemical metaphor for concurrent programming [Banatre, Le Metayer 78]
  • CHAM [Berry 85] to express the operational semantics of the Pi-Calculus
  • Membranes for modules: just like cell compartments

Hybrid continuous+discrete computations (energy + information)

  • Trend for future: more physics in informatics, more informatics in physics
slide-76
SLIDE 76

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 76

Two-stroke Engine with ATP fuel Myosin + ATP => Myosin-ATP

Myosin-ATP => Myosin + ADP

slide-77
SLIDE 77

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 77

Two-stroke Engine with ATP fuel Myosin + ATP => Myosin-ATP

Myosin-ATP => Myosin + ADP

http://www.sci.sdsu.edu/movies

slide-78
SLIDE 78

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 78

Two-stroke Engine with ATP fuel Myosin + ATP => Myosin-ATP

Myosin-ATP => Myosin + ADP

http://www.sci.sdsu.edu/movies http://www-rocq.inria.fr/sosso/icema2

slide-79
SLIDE 79

24/06/2013 François Fages - Ecole Jeunes Chercheurs - Porquerolles 79

Two-stroke Engine with ATP fuel Myosin + ATP => Myosin-ATP

Myosin-ATP => Myosin + ADP Actin-Myosin microtubule contraction controlled by MPF that phosphorylates myosin