addobservable

Add observable object to SimBiology model

Since R2020a

Syntax

obsObj = addobservable(modelObj,obsName,obsExpression)

obsObj = addobservable(modelObj,obsName,obsExpression,Name,Value)

Description

obsObj = addobservable(modelObj,obsName,obsExpression) adds an observable object to a SimBiology^® model modelObj. The inputs obsName and obsExpression are the observable object name and its expression, respectively.

obsObj = addobservable(modelObj,obsName,obsExpression,Name,Value) sets the property values of obsObj using one or more name-value pair arguments. Name is the property name and Value is the corresponding value Name must appear inside quotes. You can specify several name and value pair arguments in any order as Name1,Value1,...,NameN,ValueN. For a list of properties, see observable object properties.

Examples

Calculate Statistics After Model Simulation Using Observables

Load the Target-Mediated Drug Disposition (TMDD) model.

sbioloadproject tmdd_with_TO.sbproj

Set the target occupancy (TO) as a response.

cs = getconfigset(m1);
cs.RuntimeOptions.StatesToLog = 'TO';

Get the dosing information.

d = getdose(m1,'Daily Dose');

Scan over different dose amounts using a SimBiology.Scenarios object. To do so, first parameterize the Amount property of the dose. Then vary the corresponding parameter value using the Scenarios object.

amountParam = addparameter(m1,'AmountParam','Units',d.AmountUnits);
d.Amount = 'AmountParam';
d.Active = 1;
doseSamples = SimBiology.Scenarios('AmountParam',linspace(0,300,31));

Create a SimFunction to simulate the model. Set TO as the simulation output.

% Suppress informational warnings that are issued during simulation.
warning('off','SimBiology:SimFunction:DOSES_NOT_EMPTY');
f = createSimFunction(m1,doseSamples,'TO',d)

f = 
SimFunction

Parameters:

         Name          Value        Type            Units    
    _______________    _____    _____________    ____________

    {'AmountParam'}      1      {'parameter'}    {'nanomole'}

Observables: 

     Name         Type               Units      
    ______    _____________    _________________

    {'TO'}    {'parameter'}    {'dimensionless'}

Dosed: 

      TargetName                 TargetDimension                  Amount         AmountValue    AmountUnits 
    _______________    ___________________________________    _______________    ___________    ____________

    {'Plasma.Drug'}    {'Amount (e.g., mole or molecule)'}    {'AmountParam'}         1         {'nanomole'}


TimeUnits: day

warning('on','SimBiology:SimFunction:DOSES_NOT_EMPTY');

Simulate the model using the dose amounts generated by the Scenarios object. In this case, the object generates 31 different doses; hence the model is simulated 31 times and generates a SimData array.

doseTable = getTable(d);
sd = f(doseSamples,cs.StopTime,doseTable)

 
   SimBiology Simulation Data Array: 31-by-1
 
   ModelName:        TMDD
   Logged Data:
     Species:        0
     Compartment:    0
     Parameter:      1
     Sensitivity:    0
     Observable:     0
 

Plot the simulation results. Also add two reference lines that represent the safety and efficacy thresholds for TO. In this example, suppose that any TO value above 0.85 is unsafe, and any TO value below 0.15 has no efficacy.

h = sbioplot(sd);
time = sd(1).Time;
h.NextPlot = 'add';
safetyThreshold = plot(h,[min(time), max(time)],[0.85, 0.85],'DisplayName','Safety Threshold');
efficacyThreshold = plot(h,[min(time), max(time)],[0.15, 0.15],'DisplayName','Efficacy Threshold');

Postprocess the simulation results. Find out which dose amounts are effective, corresponding to the TO responses within the safety and efficacy thresholds. To do so, add an observable expression to the simulation data.

% Suppress informational warnings that are issued during simulation.
warning('off','SimBiology:sbservices:SB_DIMANALYSISNOTDONE_MATLABFCN_UCON');
newSD = addobservable(sd,'stat1','max(TO) < 0.85 & min(TO) > 0.15','Units','dimensionless')

 
   SimBiology Simulation Data Array: 31-by-1
 
   ModelName:        TMDD
   Logged Data:
     Species:        0
     Compartment:    0
     Parameter:      1
     Sensitivity:    0
     Observable:     1
 

The addobservable function evaluates the new observable expression for each SimData in sd and returns the evaluated results as a new SimData array, newSD, which now has the added observable (stat1).

SimBiology stores the observable results in two different properties of a SimData object. If the results are scalar-valued, they are stored in SimData.ScalarObservables. Otherwise, they are stored in SimData.VectorObservables. In this example, the stat1 observable expression is scalar-valued.

Extract the scalar observable values and plot them against the dose amounts.

scalarObs = vertcat(newSD.ScalarObservables);
doseAmounts = generate(doseSamples);
figure
plot(doseAmounts.AmountParam,scalarObs.stat1,'o','MarkerFaceColor','b')

The plot shows that dose amounts ranging from 50 to 180 nanomoles provide TO responses that lie within the target efficacy and safety thresholds.

You can update the observable expression with different threshold amounts. The function recalculates the expression and returns the results in a new SimData object array.

newSD2 = updateobservable(newSD,'stat1','max(TO) < 0.75 & min(TO) > 0.30');

Rename the observable expression. The function renames the observable, updates any expressions that reference the renamed observable (if applicable), and returns the results in a new SimData object array.

newSD3 = renameobservable(newSD2,'stat1','EffectiveDose');

Restore the warning settings.

warning('on','SimBiology:sbservices:SB_DIMANALYSISNOTDONE_MATLABFCN_UCON');

Input Arguments

modelObj — SimBiology model

SimBiology model object

SimBiology model, specified as a SimBiology model object.

obsName — Name of observable object

character vector | string

Name of the observable object, specified as a character vector or string.

The name

• Cannot contain the characters [ ], ->, or <->.

• Cannot be empty, the word time, the word null, or all whitespace.

• Must be unique in a model, meaning no observable object can have the same name as another observable, species, compartment, parameter, reaction, variant, or dose in the model.

For details, see Guidelines for Naming Model Components.

Example: 'AUC_obs'

Data Types: char | string

obsExpression — Expression of observable object

character vector | string

Expression of the observable object, specified as a character vector or string.

Example: 'trapz(time,drug)'

Data Types: char | string

Output Arguments

obsObj — Observable object

Observable object

Observable object, returned as an observable object.

Version History

Introduced in R2020a

R2024a: Having duplicate model component names is no longer allowed

• When you load a model that contains duplicate names, SimBiology automatically updates those names. SimBiology disambiguates the duplicate names by adding a suffix "_N", where N is the first positive integer that results in a unique name. If there is an existing suffix, N will be incremented from that suffix. For example, if there are two model components named x_3, the function updates one of the names to x_4. If the existing suffix has leading zeros, the function omits the zeros in the new name. For instance, if x_003 is a duplicate name, it gets renamed to x_4. However, the function assumes that names with leading zeros and without leading zeros are different. For instance, x_005 and x_5 are considered to be different names.

• SimBiology issues an error if multiple model components (model, compartment, species, parameter, reaction, rule, event, observable, dose, and variant) have the same name. Within a single model, these components are required to have unique names even when they are of different types with the following two exceptions:

  • Species in different compartments can have the same name.

  • Parameters can have the same name if they are scoped to different parents. Specifically, you can use the same name for a model-scoped parameter and reaction-scoped parameters, where each reaction-scoped parameter belongs to a different reaction.

  For details on how to reference model component names in expressions, see Guidelines for Referencing Names in Expressions.

R2022b: Having duplicate model component names issues a warning

SimBiology issues a warning if multiple model components have the same name.

R2022a: Having duplicate model component names will not be allowed in a future release

SimBiology will not allow you to have duplicate names for model components within a model.