Transcript Lesson: Cobra Toolbox BIE 5500/6500 Utah State University

Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Learning Objectives
Each student should be able to:
• Explain the purpose of the Cobra Toolbox,
• Demonstrate basic operation of the Cobra Toolbox,
• Explain the capabilities of BIGG Database,
• Explain the capabilities of Pathway Pioneer,
• Explain the capabilities of Paint4Net.
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
COBRA TOOLBOX OVERVIEW
Matlab Cobra Toolbox
SBML, Excel
Load Models
Output Maps
• Flux Optimization
• Robustness Analysis
Numerical Output
• Phenotype Phase Plane Analysis
• Flux Variability Analysis
M-Files
Matlab Code
Graphical Output
• Gene Additions & Knockouts
• Production Envelopes
Save Models
• More Tools
Schellenberger J, Que R, Fleming RMT, Thiele I, Orth JD, Feist AM, Zielinski DC, Bordbar A, Lewis NE, Rahmanian S, Kang J, Hyduke DR, Palsson BØ.
2011 Quantitative prediction of cellular metabolism with constraint-based models: the COBRA Toolbox v2.0. Nature Protocols 6:1290-1307.
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
COBRA TOOLBOX
INSTALLATION
http://sourceforge.net/projects/opencobra/files/cobra/
http://opencobra.sourceforge.net/openCOBRA/Install.html
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Matlab
GUI
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
Starting Matlab
Start Matlab
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Initializing the
Cobra Toolbox
Run initCobraToolbox.m
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Command
Reactions
Metabolites
Loading a
Cobra-based
Model
Loaded Model
S matrix
Lower Bounds
Cobra Model
Structure
Upper Bounds
Objective
Function
Load Model: model=readCbModel('ecoli_textbook');
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
Double click on
Model to open
window with the
model variables
Cobra Model
Structure
Utah State University
© 2015 H. Scott Hinton
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
MODEL STRUCTURE USED BY THE COBRA TOOLBOX
• rxns: A list of all of the reaction abbreviations in the same order they appear in the stoichiometric matrix
• mets: A list of all of the metabolite abbreviations in the model in the same order they appear in the stoichiometric matrix
• S: The stoichiometric matrix in sparse format
• lb: The lower bound corresponding to each reaction, in order
• ub: The upper bound corresponding to each reaction, in order
• c: The relative weight of each reaction in the objective function—often a single one in the position corresponding to the biomass
reaction and zeros elsewhere
• subSystem: The metabolic subsystem for each reaction
• rules: Boolean rules for each reaction describing the gene-reaction relationship. For example ‘gene1 and gene2’ indicate that the two
gene products are part of a enzyme comples whereas ‘gene1 or gene2’ indicate that the two gene products are isozymes that catalyze
the same reaction.
• genes: The gene names of all the genes included in the model
• rxnGeneMat: A matrix with as many rows as there are reactions in the model and as many columns as there are genes in the model. The
ith row and jth column contains a one if the jth gene in genes is associated with the ith reaction in rxns and zero otherwise.
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
DRAWING MAPS
• Maps for a variety of metabolic pathways are available
for many of the models hosted in the BiGG
knowledgebase (http://bigg.ucsd.edu).
• Read the map into Cobra
map=readCbMap('ecoli_Textbook_ExportMap');
• Cobra saves a map into a file named “target.svg” that
will be written into the current directory.
drawCbMap(map);
target.svg
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
MODEL SPREADSHEET
ecoli_core_model.xls [writeCbModel(model, ‘xls’)]
abbreviation
ACALD
ACALDt
ACKr
ACONTa
ACONTb
ACt2r
ADK1
AKGDH
AKGt2r
ALCD2x
ATPM
ATPS4r
CO2t
CS
officialName
acetaldehyde dehydrogenase (acetylating)
acetaldehyde reversible transport
acetate kinase
aconitase (half-reaction A, Citrate hydro-lyase)
aconitase (half-reaction B, Isocitrate hydro-lyase)
acetate reversible transport via proton symport
adenylate kinase
2-Oxogluterate dehydrogenase
2-oxoglutarate reversible transport via symport
alcohol dehydrogenase (ethanol)
ATP maintenance requirement
ATP synthase (four protons for one ATP)
CO2 transporter via diffusion
citrate synthase
equation
[c] : acald + coa + nad <==> accoa + h + nadh
acald[e] <==> acald[c]
[c] : ac + atp <==> actp + adp
[c] : cit <==> acon-C + h2o
[c] : acon-C + h2o <==> icit
ac[e] + h[e] <==> ac[c] + h[c]
[c] : amp + atp <==> (2) adp
[c] : akg + coa + nad --> co2 + nadh + succoa
akg[e] + h[e] <==> akg[c] + h[c]
[c] : etoh + nad <==> acald + h + nadh
[c] : atp + h2o --> adp + h + pi
adp[c] + (4) h[e] + pi[c] <==> atp[c] + (3) h[c] + h2o[c]
co2[e] <==> co2[c]
[c] : accoa + h2o + oaa --> cit + coa + h
subSystem
Pyruvate Metabolism
Transport, Extracellular
Pyruvate Metabolism
Citric Acid Cycle
Citric Acid Cycle
Transport, Extracellular
Oxidative Phosphorylation
Citric Acid Cycle
Transport, Extracellular
Pyruvate Metabolism
Oxidative Phosphorylation
Oxidative Phosphorylation
Transport, Extracellular
Citric Acid Cycle
CYTBD
D_LACt2
ENO
ETOHt2r
EX_ac(e)
EX_acald(e)
EX_akg(e)
abbreviation
ACALD
ACALDt
ACKr
ACONTa
ACONTb
cytochrome oxidase bd (ubiquinol-8: 2 protons)
D-lactate transport via proton symport
enolase
ethanol reversible transport via proton symport
Acetate exchange
Acetaldehyde exchange
2-Oxoglutarate exchange
officialName
acetaldehyde dehydrogenase (acetylating)
acetaldehyde reversible transport
acetate kinase
aconitase (half-reaction A, Citrate hydro-lyase)
aconitase (half-reaction B, Isocitrate hydro-lyase)
(2) h[c] + (0.5) o2[c] + q8h2[c] --> (2) h[e] + h2o[c] + q8[c]
h[e] + lac-D[e] <==> h[c] + lac-D[c]
[c] : 2pg <==> h2o + pep
etoh[e] + h[e] <==> etoh[c] + h[c]
[e] : ac <==>
[e] : acald <==>
[e] : akg <==>
equation
[c] : acald + coa + nad <==> accoa + h + nadh
acald[e] <==> acald[c]
[c] : ac + atp <==> actp + adp
[c] : cit <==> acon-C + h2o
[c] : acon-C + h2o <==> icit
Oxidative Phosphorylation
Transport, Extracellular
Glycolysis/Gluconeogenesis
Transport, Extracellular
Exchange
Exchange
Exchange
subSystem
Pyruvate Metabolism
Transport, Extracellular
Pyruvate Metabolism
Citric Acid Cycle
Citric Acid Cycle
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
MODIFYING COBRA TOOLBOX MODELS
• The cobra model can be modified to simulate different conditions by altering reaction flux
rates. This includes, adding and/or removing reactions or changing the model objective
function.
• Altering reaction bounds (adding and/or removing reactions)
• model = changeRxnBounds(model, rxnNameList, value, boundType);
• rxnNameList is a cell array of reaction IDs corresponding to reaction IDs in model.rxns; value is a floating
point number representing the bound of the flux rate through the reaction; boundType specifies which
bounds to change for the reactions and can take values of ‘l’, ‘u’, or ‘b’ for lower, upper, or both, respectively.
• Add reaction
• [model] = addReaction(model, rxnName, metaboliteList, stoichCoeffList, [revFlag], [lowerBound],
[upperBound], [objCoeff], [subsystem], [grRule], [geneNameList], [systNameList],
[checkDuplicate]);
• Remove reaction
• [model] = removeRxns(model, rxnRemoveList)
• Change the objective function (typically growth)
• model = changeObjective(model, rxnNameList, [objectiveCoeff]);
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Structure used
to store solution
Solving
for Fluxes
Solve for Fluxes
solution = optimizeCbModel(model,'max');
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Flux Vector
Objective Function Value
Shadow Prices
Structure of
Optimization
Solution
Utah State University
Reduced Costs
Solver
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
Printing the
Optimized Flux
Solutions
© 2015 H. Scott Hinton
Flux Passing through the
Reaction (mmols gDW-1h-1)
Reaction Name
Printing Flux Solutions
printFluxVector(model, solution.x, true)
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Print Flux Values
printFluxVector(model, solution.x, true)
ACONTa 6.00725
ACONTb 6.00725
AKGDH
5.06438
ATPM
8.39
ATPS4r 45.514
Biomass_...0.873922
CO2t
-22.8098
CS
6.00725
CYTBD
43.599
ENO
14.7161
EX_co2(e) 22.8098
EX_glc(e) -10
EX_h2o(e) 29.1758
EX_h(e) 17.5309
EX_nh4(e) -4.76532
EX_o2(e) -21.7995
EX_pi(e) -3.2149
Utah State University
Growth
Rate
Inputs & Outputs
(Exchange
Reactions)
FBA
FUM
G6PDH2r
GAPD
GLCpts
GLNS
GLUDy
GND
H2Ot
ICDHyr
MDH
NADH16
NH4t
O2t
PDH
PFK
PGI
7.47738
5.06438
4.95998
16.0235
10
0.223462
-4.54186
4.95998
-29.1758
6.00725
5.06438
38.5346
4.76532
21.7995
9.28253
7.47738
4.86086
BIE 5500/6500
PGK
-16.0235
PGL
4.95998
PGM
-14.7161
PIt2r
3.2149
PPC
2.50431
PYK
1.75818
RPE
2.67848
RPI
-2.2815
SUCDi
5.06438
SUCOAS -5.06438
TALA
1.49698
TKT1
1.49698
TKT2
1.1815
TPI
7.47738
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
printFluxVector
printFluxVector(model, solution.x)
ACALD
ACALDt
ACKr
ACONTa
ACONTb
ACt2r
ADK1
AKGDH
AKGt2r
ALCD2x
ATPM
ATPS4r
Biomass...
CO2t
CS
CYTBD
D_LACt2
ENO
ETOHt2r
EX_ac(e)
EX_acald(e)
EX_akg(e)
EX_co2(e)
EX_etoh(e)
EX_for(e)
EX_fru(e)
EX_fum(e)
EX_glc(e)
EX_gln_L(e)
-14.6749
0
-15.1732
0.507693
0.507693
-15.1732
0
0
0
-14.6749
8.39
-11.1879
0.470565
0.840759
0.507693
0
0
35.0451
-14.6749
15.1732
-0
-0
-0.840759
14.6749
32.1194
-0
-0
-18.5
-0
Utah State University
printFluxVector(model, solution.x, true)
ACALD
ACKr
ACONTa
ACONTb
ACt2r
ALCD2x
ATPM
ATPS4r
Biomass...
CO2t
CS
ENO
ETOHt2r
EX_ac(e)
EX_co2(e)
EX_etoh(e)
EX_for(e)
EX_glc(e)
EX_h2o(e)
EX_h(e)
EX_nh4(e)
EX_pi(e)
-14.6749
-15.1732
0.507693
0.507693
-15.1732
-14.6749
8.39
-11.1879
0.470565
0.840759
0.507693
35.0451
-14.6749
15.1732
-0.840759
14.6749
32.1194
-18.5
-12.0879
56.7321
-2.5659
-1.73107
printFluxVector(model, solution.x, true,true)
Biomass...
EX_ac(e)
EX_co2(e)
EX_etoh(e)
EX_for(e)
EX_glc(e)
EX_h2o(e)
EX_h(e)
EX_nh4(e)
EX_pi(e)
0.470565
15.1732
-0.840759
14.6749
32.1194
-18.5
-12.0879
56.7321
-2.5659
-1.73107
Only prints Exchange
Reactions that are nonzero
Only prints Reactions
that are nonzero
http://opencobra.sourceforge.net/openCOBRA/opencobra_documentation/cobra_toolbox_2/index.html
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
DRAWING FLUX VALUES
ONTO A MAP
drawFlux(map, model, solution.x)
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
MATLAB
EDITOR
Clear the Workspace
Run m-file
Supports multiple
m-file windows
• The Matlab editor
allows m-files to be
created
• m-files can contain a
list of Matlab
commands to be
executed sequentially.
Comments
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Aerobic E.coli Growth on Glucose
("What is flux balance analysis? - Supplementary tutorial“)
Set the maximum glucose uptake rate to 18.5 mmol gDW-1 hr-1 (millimoles per gram dry cell weight per hour, the default flux units used in the COBRA
Toolbox), enter:
model = changeRxnBounds(model,'EX_glc(e)',-18.5,'l');
This changes the lower bound ('l') of the glucose exchange reaction to -18.5, a biologically realistic uptake rate (note that the import of a metabolite
is listed as a negative flux). To allow unlimited oxygen uptake, enter:
model = changeRxnBounds(model,'EX_o2(e)',-1000,'l');
By setting the lower bound of the oxygen uptake reaction to such a large number, it is practically
unbounded. Set the biomass reaction is set as the objective function, enter:
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2');
To perform FBA with maximization of the biomass reaction as the objective, enter:
FBAsolution = optimizeCbModel(model,'max');
FBAsolution.f then gives the value of the objective function (Z) as 1.6531 (the model predicts a growth rate of 1.6531 hr -1). The flux distribution
vector FBAsolution.x shows that there is high flux in the glycolysis, pentose phosphate, TCA cycle, and oxidative phosphorylation pathways, with no
secreted organic by-products. See AerobicGlucoseBioMass.m
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
AerobicGlucoseBioMass.m
clear; % Clear Workspace
model=readCbModel('ecoli_textbook'); % Read textbook model
model = changeRxnBounds(model,'EX_glc(e)',-18.5,'l'); % Set lower bound of glucose
model = changeRxnBounds(model,'EX_o2(e)',-1000,'l'); % Set lower bound of oxygen
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2'); % Set objective function (Biomass)
FBAsolution = optimizeCbModel(model,'max‘,0,0); % Find optimized flux values
map=readCbMap('ecoli_Textbook_ExportMap'); % Input ecoli textbook map template
options.zeroFluxWidth = 0.1;
options.rxnDirMultiplier = 10;
drawFlux(map, model, FBAsolution.x, options); % Draw Map
printFluxVector(model, FBAsolution.x, true); % Print flux values
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
EX_glc(e)
Aerobic Growth on Glucose
Exchange Reactions
Biomass
1.65
EX_co2(e)
40.6527
EX_glc(e)
-18.5
EX_h2o(e)
52.6943
EX_h(e)
33.1606
EX_nh4(e)
-9.01387
EX_o2(e)
-38.7416
EX_pi(e)
-6.08116
EX_o2(e)
EX_h(e)
AerobicGlucoseBioMass.m
EX_pi(e) EX_h2o(e) EX_co2(e) EX_nh4(e)
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Anaerobic E.coli Growth on Glucose
("What is flux balance analysis? - Supplementary tutorial“)
Set the maximum glucose uptake rate to 18.5 mmol gDW-1 hr-1 (millimoles per gram dry cell weight per hour, the default flux units used in
the COBRA Toolbox), enter:
model = changeRxnBounds(model,'EX_glc(e)',-18.5,'l');
This changes the lower bound ('l') of the glucose exchange reaction to -18.5, a biologically realistic uptake rate (note that the import of a
metabolite is listed as a negative flux). To prevent oxygen uptake, enter:
model = changeRxnBounds(model,'EX_o2(e)',0,'l');
Anaerobic operation is achieved by setting the lower bound of the oxygen uptake reaction to zero.
Set the biomass reaction is set as the objective function, enter:
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2‘,0,0);
To perform FBA with maximization of the biomass reaction as the objective, enter:
FBAsolution = optimizeCbModel(model,'max');
FBAsolution.f then gives the value of the objective function (Z) as 0.4706 (the model predicts a growth rate of 0.4706 hr-1). The flux
distribution vector FBAsolution.x shows that oxidative phosphorylation is not used in these conditions, and that acetate, formate, and
ethanol are produced by fermentation pathways. See AnaerobicGlucoseBioMass.m
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Anaerobic Growth on Glucose
Exchange Reactions
Biomass
0.470565
EX_ac(e)
15.1732
EX_co2(e)
-0.840759
AnaerobicGlucoseBioMass.m
EX_etoh(e) 14.6749
EX_for(e)
32.1194
EX_glc(e)
-18.5
EX_h2o(e)
-12.0879
EX_h(e)
56.7321
EX_nh4(e)
-2.5659
EX_pi(e)
-1.73107
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
AEROBIC vs. ANAEROBIC GROWTH
Orth, J. D., I. Thiele, et al. (2010). "What is flux balance analysis?" Nature biotechnology 28(3): 245-248.
a.
b.
Aerobic Growth
Utah State University
Anaerobic Growth
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Growth on Alternate Substrates
("What is flux balance analysis? - Supplementary tutorial“)
The core E. coli model contains exchange reactions for 13 different organic compounds, each of which can be used as the
sole carbon source under aerobic conditions. For example, to simulate growth on succinate instead of glucose
model = changeRxnBounds(model,'EX_glc(e)',0,'l'); % Required if glucose uptake is zero
model = changeRxnBounds(model,'EX_succ(e)',-20,'l');
FBAsolution = optimizeCbModel(model,'max');
The growth rate, given by FBAsolution.f, will be 0.8401 hr-1
Growth can also be simulated under anaerobic conditions with any substrate by using changeRxnBounds to set the lower
bound of the oxygen exchange reaction (EX_o2(e)) to 0 mmol gDW-1 hr-1, so no oxygen can enter the system. When this
constraint is applied and succinate is the only organic substrate, optimizeCbModel returns a growth rate of 0 hr-1, and
does not calculate a flux vector v (depending on which linear programming solver is used with the COBRA Toolbox, a
growth rate may not be calculated at all). In this case, FBA predicts that growth is not possible on succinate under
anaerobic conditions. Because the maximum amount of ATP that can be produced from this amount of succinate is less
than the minimum bound of 8.39 mmol gDW-1 hr-1 of the ATP maintenance reaction, ATPM, there is no feasible
solution.organic by-products.
See AerobicSuccinateBioMass.m and AnaerobicSuccinateBioMass.m
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Growth on Succinate
AerobicSuccinateBioMass.m
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Substrate Maximum Growth Rate
Substrate
The core E. coli model
contains exchange reactions
for 13 different organic
compounds, each of which
can be used as the sole
carbon source under aerobic
or anaerobic conditions.
Aerobic (hr-1)
Anaerobic (hr-1)
acetate
0.3893
0
acetaldehyde
0.6073
0
2-oxoglutarate
1.0982
0
ethanol
0.6996
0
D-fructose
1.7906
0.5163
fumarate
0.7865
0
D-glucose
1.7906
0.5163
L-glutamine
1.1636
0
L-glutamate
1.2425
0
D-lactate
0.7403
0
L-malate
0.7865
0
pyruvate
0.6221
0.0655
succinate
0.8401
0
("What is flux balance analysis? - Supplementary tutorial“)
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
( http://bigg.ucsd.edu/bigg/home.pl )
Full E.coli model “ecoli_iaf1260.xml”
Schellenberger, J., J. O. Park, et al. (2010). "BiGG: a Biochemical Genetic and Genomic knowledgebase of large scale metabolic reconstructions." BMC Bioinformatics 11: 213.
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Pathway Pioneer
http://pathwaypioneer.com/
• Can be used with multiple models
• Visually centered
• Intuitive to use
• Editing capability
• Embedded search engine
• Includes chemical information
• One-click operations
• Rapid explorations
• Save and share designs/results
Utah State University
Requires Firefox or Chrome
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Visualization Tools
BiliverdinOptimize_Visualize.m
BIGG Database
Paint4Net Tool
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Paint4Net
http://www.biosystems.lv/index.php?option=com_content&view=category&layout=blog&id=65&Itemid=86
• Developed by Andrejs Kostromins
• Paint4Net is the COBRA Toolbox extension for visualization of
constraints-based reconstruction and analysis (COBRA) models
and reconstructions in the MATLAB environment.
• Uses the Bioinformatics toolbox to visualize COBRA models and
reconstructions as a hypergraph.
• Paint4Net contains two main commands:
• draw_by_rxn
• For visualization of all or a part of a COBRA model by
specified list of reactions.
• draw_by_met
• For visualization of the connectivity of a particular
metabolite with other metabolites through reactions of a
COBRA model
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
draw_by_rxn
• Rectangles represent reactions;
• Numbers in rectangles represent flux rate through
reaction.
• Red rectangles represent reactions with only one input
or output flux (signaling a potential dead reaction);
• Ellipses represent metabolites;
• Red ellipses represent dead end metabolites;
• Grey edges represent zero-rate fluxes;
• Green edges represent positive-rate (forward) fluxes;
• Blue edges represent negative-rate (backward) fluxes.
• The thickness of the edges is calculated as percentage
assuming the maximum rate of flux in the model
corresponds to 100%.
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Plotting Textbook Model
VisualizeFlux_Textbook.m
clear;
model=readCbModel('ecoli_textbook');
model = changeRxnBounds(model,'EX_glc(e)',-5,'l');
model = changeRxnBounds(model,'EX_o2(e)',-20,'l');
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2');
FBAsolution = optimizeCbModel(model,'max‘,0,0);
cofactors =
{'cmp[c]','ctp[c]','acp[c]','ACP[c]','malACP[c]','amp[c]','atp[c]','adp[c]','pi[c]','ppi[c]','nad[c]','nadh[c]','nadph[c]','nadp[c
]','h[c]','h2o[c]','co2[c]'};
% Plot includes cofactors
[involvedMets,deadEnds]= draw_by_rxn (model,model.rxns,true,'struc',{''},{''},FBAsolution.x);
% Plot removes cofactors
[involvedMets,deadEnds]= draw_by_rxn (model,model.rxns,true,'struc',{''},cofactors,FBAsolution.x);
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Textbook Model With Cofactors
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Textbook Model Without Cofactors
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Plotting Active Reactions in Core Model
VisualizeFlux_core.m
clear;
model=readCbModel('ecoli_textbook');
model = changeRxnBounds(model,'EX_glc(e)',-5,'l');
model = changeRxnBounds(model,'EX_o2(e)',-20,'l');
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2');
FBAsolution = optimizeCbModel(model,'max‘,0,0);
rxnID = findRxnIDs(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2');
m = 0;
[n,nLab] = size(model.rxns);
for i=1:n
if(i~=rxnID)
if(FBAsolution.x(i) ~= 0)
m = m+1;
fluxReactions(m) = model.rxns(i);
end
end
end
cofactors = {'cmp[c]','ctp[c]','acp[c]','ACP[c]','malACP[c]','amp[c]','atp[c]','adp[c]','pi[c]',
ppi[c]','nad[c]','nadh[c]','nadph[c]','nadp[c]','h[c]','h2o[c]','co2[c]'};
[involvedMets,deadEnds]= draw_by_rxn (model,fluxReactions,true,'struc',{''},cofactors,FBAsolution.x);
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Plot of the Active Reactions
in Core Model
VisualizeFlux_core.m
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Model Subsystems (Reactions)
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Plotting Model Subsystems
Paint4Net_core_subSystem.m
clear;
model=readCbModel('ecoli_textbook');
model = changeRxnBounds(model,'EX_glc(e)',-5,'l');
model = changeRxnBounds(model,'EX_o2(e)',-20,'l');
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2');
FBAsolution = optimizeCbModel(model,'max');
% Identify cofactors
cofactors = {'atp[c]','adp[c]','pi[c]','nad[c]','nadh[c]','nadph[c]','nadp[c]','h[c]','h2o[c]'};
%Extract & plot single subsystem reactions (Citric Acid Cycle)
fluxReactions = model.rxns(ismember(model.subSystems,'Citric Acid Cycle'));
[involvedMets,deadEnds]= draw_by_rxn (model,fluxReactions,true,'struc',{''},{''},FBAsolution.x);
% Extract multiple subsystem reactions
includedSubSystems = {'Citric Acid Cycle','Pyruvate Metabolism','Oxidative
Phosphorylation','Glycolysis/Gluconeogenesis','Pentose Phosphate Pathway','Glutamate Metabolism'};
fluxReactions = model.rxns(ismember(model.subSystems,includedSubSystems));
% Plot multiple subsystem
[involvedMets,deadEnds]= draw_by_rxn (model,fluxReactions,true,'struc',{''},cofactors,FBAsolution.x);
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Paint4Net Plots
Citric Acid Cycle
Utah State University
BIE 5500/6500
Multiple Subsystems
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
draw_by_met
• Rectangles represent reactions;
• Numbers in rectangles represent flux rate through reaction.
• Red rectangles represent reactions with only one input or output
flux (signaling a potential dead reaction);
• Ellipses represent metabolites;
• Red ellipses represent dead end metabolites;
• Grey edges represent zero-rate fluxes;
• Green edges represent positive-rate (forward) fluxes;
• Blue edges represent negative-rate (backward) fluxes.
• The thickness of the edges is calculated as percentage assuming
the maximum rate of flux in the model corresponds to 100%.
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Plotting ‘etoh[c]’ Connectivity
VisualizeMets_Textbook.m
clear;
model=readCbModel('ecoli_textbook');
model = changeRxnBounds(model,'EX_glc(e)',-5,'l');
model = changeRxnBounds(model,'EX_o2(e)',-20,'l');
model = changeObjective(model,'Biomass_Ecoli_core_N(w/GAM)_Nmet2');
FBAsolution = optimizeCbModel(model,'max');
cofactors =
{'cmp[c]','ctp[c]','acp[c]','ACP[c]','malACP[c]','amp[c]','atp[c]','adp[c]','pi[c]','ppi[c]','nad[c]','nadh[c
]','nadph[c]','nadp[c]','h[c]','h2o[c]','co2[c]'};
% Plot connectivity to 'etoh[c]', include cofactors. Radius = 1
[invovledRxns,involvedMets,deadEnds]= draw_by_met (model,{'etoh[c]'},true,1,'struc',{''},FBAsolution.x);
% Plot connectivity to 'etoh[c]', remove cofactors. Radius = 1
[invovledRxns,involvedMets,deadEnds]= draw_by_met (model,{'etoh[c]'},true,1,'struc',cofactors,FBAsolution.x);
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
‘etoh[c]’ Connectivity (Radius = 1)
VisualizeMets_Textbook.m
With Cofactors
Utah State University
Without Cofactors
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
‘etoh[c]’ Connectivity (Radius = 2)
VisualizeMets_Textbook.m
Without Cofactors
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox
Constraint-based Metabolic Reconstructions & Analysis
© 2015 H. Scott Hinton
Cobra Toolbox
• Cobra Toolbox Overview
• Cobra Toolbox Fundamentals
• Cobra Examples
• BIGG Database & Maps
• Pathway Pioneer
• Paint4Net
Utah State University
BIE 5500/6500
Lesson: Cobra Toolbox