UseCase4_1 : Warfarin population pharmacokinetic model for infusion and oral administration routes

Variant on UseCase4 : Using COMPARTMENTS to specify PK structural model

This example script is intended to illustrate how to use the 'ddmore' R package to perform a M&S workflow using the DDMoRe Standalone Execution Environment (SEE).

The following steps are implemented in this workflow:

To run a task, select with the cursor any code lines you wish to execute and press CTRL+R+R in your keyboard. An HTML file containing the commands in this file and associated output will be provided to allow the user to compare the results

Initialisation

Clear workspace and set working directory under 'UsesCasesDemo' project

rm(list=ls(all=FALSE))
mydir <- file.path(Sys.getenv("MDLIDE_WORKSPACE_HOME"),"UseCasesDemo")
setwd(mydir)

Create a working directory under 'models' folder where results are stored

uc<-"UseCase4_1"
datafile <- "warfarin_infusion_oral.csv"
mdlfile <- paste0(uc,".mdl")

Create a new folder to be used as working directory and where results will be stored

wd <- file.path(mydir,uc)
dir.create(wd)
## Warning in dir.create(wd):
## 'D:\SEE-Prod5_RC4\MDL_IDE\workspace\UseCasesDemo\UseCase4_1' already
## exists

Copy the dataset and the .mdl file available under “models” into the working directory

file.copy(file.path(mydir,"models", datafile),wd)
## [1] TRUE
file.copy(file.path(mydir,"models",mdlfile),wd)
## [1] TRUE

Set the working directory.

setwd(file.path(mydir,uc))

The working directory needs to be set differently when knitr R package is used to spin the file

library(knitr)
opts_knit$set(root.dir = file.path(mydir,uc)) 

List files available in working directory

list.files()
## [1] "UseCase4_1.mdl"             "warfarin_infusion_oral.csv"

Model Development

ESTIMATE model parameters using Monolix

mlx <- estimate(mdlfile, target="MONOLIX", subfolder="Monolix")
## -- Wed Aug 17 18:37:03 2016
## New
## Submitted
## Job b394f4c7-fa5c-4054-bf0b-b4ff08a3c0ed progress:
## Running [ .......... ]
## Importing Results
## Copying the result data back to the local machine for job ID b394f4c7-fa5c-4054-bf0b-b4ff08a3c0ed...
## From C:\Users\zparra\AppData\Local\Temp\RtmpSKgKTU\DDMORE.job1dec334ee8d to D:/SEE-Prod5_RC4/MDL_IDE/workspace/UseCasesDemo/UseCase4_1/Monolix
## Done.
## 
## 
## The following main elements were parsed successfully:
##   ToolSettings
##   RawResults
##   Estimation::PopulationEstimates::MLE
##   Estimation::PrecisionPopulationEstimates::MLE
##   Estimation::IndividualEstimates::Estimates
##   Estimation::IndividualEstimates::RandomEffects
##   Estimation::Residuals::ResidualTable
##   Estimation::Predictions
##   Estimation::OFMeasures::IndividualContribToLL
##   Estimation::OFMeasures::InformationCriteria
##   Estimation::OFMeasures::LogLikelihood
## 
## Completed
## -- Wed Aug 17 18:40:32 2016
slotNames(mlx)
## [1] "ToolSettings"     "RawResults"       "TaskInformation" 
## [4] "Estimation"       "ModelDiagnostic"  "Simulation"      
## [7] "OptimalDesign"    ".pathToSourceXML"

The ddmore “LoadSOObj” reads and parsed existing Standardise Output Objects

# mlx <- LoadSOObject("Monolix/UseCase4_1.SO.xml")

The ddmore function “getPopulationParameters” extracts the Population Parameter values from an R object of (S4) class “StandardOutputObject” and returns the estimates. See documentation for getPopulationParameters to see other arguments and settings for this function.

parameters_mlx <- getPopulationParameters(mlx, what="estimates")
print(parameters_mlx)
## $MLE
##      POP_V  BETA_V_WT     POP_KA     POP_CL BETA_CL_WT   POP_TLAG 
##    7.96620    1.00000    0.35710    0.10026    0.75000    1.00150 
##  POP_FORAL      PPV_V     PPV_KA     PPV_CL   PPV_TLAG  PPV_FORAL 
##    0.98421    0.12539    0.06625    0.12171    0.10000    0.41025 
##  CORR_CL_V    RUV_ADD   RUV_PROP 
##    0.02292    0.00000    0.12309
print(getPopulationParameters(mlx, what="precisions"))
## $MLE
##     Parameter     MLE      SE        RSE
## 1  BETA_CL_WT 0.75000 0.00000       0.00
## 2   BETA_V_WT 1.00000 0.00000       0.00
## 3   CORR_CL_V 0.02292 0.18890     824.17
## 4      POP_CL 0.10026 0.00225       2.25
## 5   POP_FORAL 0.98421 0.01043       1.06
## 6      POP_KA 0.35710 0.01291       3.61
## 7    POP_TLAG 1.00150 0.03110       3.11
## 8       POP_V 7.96620 0.18562       2.33
## 9      PPV_CL 0.12171 0.01611      13.24
## 10  PPV_FORAL 0.41025 3.75141     914.42
## 11     PPV_KA 0.06625 0.05279      79.69
## 12   PPV_TLAG 0.10000 0.00000       0.00
## 13      PPV_V 0.12539 0.01692      13.49
## 14    RUV_ADD 0.00000 0.02155 1554819.49
## 15   RUV_PROP 0.12309 0.00458       3.72
print(getEstimationInfo(mlx))
## $OFMeasures
## $OFMeasures$LogLikelihood
## $OFMeasures$LogLikelihood[[1]]
## [1] -1337.88
## 
## 
## $OFMeasures$IndividualContribToLL
##    Subject ICtoLL
## 1        1 -34.25
## 2        2 -36.07
## 3        3 -27.39
## 4        4 -46.31
## 5        5 -43.62
## 6        6 -46.02
## 7        7 -37.17
## 8        8 -43.31
## 9        9 -34.61
## 10      10 -59.79
## 11      11 -42.53
## 12      12 -41.62
## 13      13 -49.72
## 14      14 -44.59
## 15      15 -48.87
## 16      16 -42.59
## 17      17 -39.70
## 18      18 -45.33
## 19      19 -40.84
## 20      20 -46.72
## 21      21 -40.62
## 22      22 -36.46
## 23      23 -51.25
## 24      24 -45.10
## 25      25 -43.12
## 26      26 -44.94
## 27      27 -40.12
## 28      28 -37.96
## 29      29 -33.18
## 30      30 -33.74
## 31      31 -45.96
## 32      32 -34.40
## 
## $OFMeasures$InformationCriteria
## $OFMeasures$InformationCriteria$AIC
## [1] 2699.76
## 
## $OFMeasures$InformationCriteria$BIC
## [1] 2717.35
## 
## 
## 
## $Messages
## list()

Perform model diagnostics for the base model using Xpose functions (graphs are exported to PDF)

Use 'ddmore' function as.xpdb() to create an Xpose database object from the Standardised Output object, regardless of target software used for estimation.

mlx.xpdb<-as.xpdb(mlx,datafile)
## 
## Removed dose rows in rawData slot of SO to enable merge with Predictions data.

We can then call Xpose functions referencing this mlx.xpdb object as the input. Perform some basic goodness of fit

print(basic.gof(mlx.xpdb))

plot of chunk unnamed-chunk-10

print(ind.plots(mlx.xpdb))

plot of chunk unnamed-chunk-10 plot of chunk unnamed-chunk-10

Export results to PDF file

pdf("GOF_MLX.pdf")
print(basic.gof(mlx.xpdb))
print(ind.plots(mlx.xpdb))
dev.off()
## rj.GD 
##     2

Change estimation method to FOCEI (for speed)

MDL Objects can be manipulated from R to change for example the estimation algorithm

myTaskProperties <- getTaskPropertiesObjects(mdlfile)[[1]]
myNewTaskProperties <- myTaskProperties
myNewTaskProperties@ESTIMATE$algo <- "focei"

Assembling the new MOG. Note that we reuse the data and model from the previous run.

myNewerMOG <- createMogObj(dataObj = getDataObjects(mdlfile)[[1]], 
        parObj = getParameterObjects(mdlfile)[[1]], 
        mdlObj = getModelObjects(mdlfile)[[1]], 
        taskObj = myNewTaskProperties)

We can then write the MOG back out to an .mdl file.

mdlfile.FOCEI <- paste0(uc,"_FOCEI.mdl")
writeMogObj(myNewerMOG,mdlfile.FOCEI)

Test estimation using this new MOG.

NM.FOCEI <- estimate(mdlfile.FOCEI, target="PsN", subfolder="NONMEM_FOCEI")
## -- Wed Aug 17 18:42:03 2016
## New
## Submitted
## Job 8df22d6b-1cfe-4dff-bf91-ec9cad24fe1c progress:
## Running [ .... ]
## Importing Results
## Copying the result data back to the local machine for job ID 8df22d6b-1cfe-4dff-bf91-ec9cad24fe1c...
## From C:\Users\zparra\AppData\Local\Temp\RtmpSKgKTU\DDMORE.job1dec533c4eb1 to D:/SEE-Prod5_RC4/MDL_IDE/workspace/UseCasesDemo/UseCase4_1/NONMEM_FOCEI
## Done.
## 
## 
## The following main elements were parsed successfully:
##   RawResults
##   Estimation::PopulationEstimates::MLE
##   Estimation::IndividualEstimates::Estimates
##   Estimation::IndividualEstimates::RandomEffects
##   Estimation::Residuals::ResidualTable
##   Estimation::Predictions
##   Estimation::OFMeasures::Deviance
## 
## The following WARNINGs were raised during the job execution:
##  estimation_successful: 0
## 
## The following MESSAGEs were raised during the job execution:
##  covariance_step_run: 0
##  rounding_errors: 0
##  hessian_reset: 0
##  zero_gradients: 0
##  final_zero_gradients: 0
##  estimate_near_boundary: 0
##  s_matrix_singular: 0
##  nmoutput2so_version: This SOBlock was created with nmoutput2so version 4.5.27
## 
## Completed
## -- Wed Aug 17 18:43:32 2016

Load previous results

# NM.FOCEI <- LoadSOObject("NONMEM_FOCEI/UseCase4_1_FOCEI.SO.xml")

Results from NONMEM should be comparable to previous results

print(getPopulationParameters(NM.FOCEI,  what="estimates"))
## $MLE
##     POP_CL      POP_V     POP_KA   POP_TLAG  POP_FORAL   RUV_PROP 
##  0.1001090  8.0906700  0.3771930  1.1207400  0.9999930  0.1300870 
##    RUV_ADD BETA_CL_WT  BETA_V_WT     PPV_CL  CORR_CL_V      PPV_V 
##  0.0191873  0.7500000  1.0000000  0.1319380  0.0113956  0.1197450 
##     PPV_KA   PPV_TLAG  PPV_FORAL 
##  0.0350171  0.1000000  0.1011060
print(parameters_mlx)
## $MLE
##      POP_V  BETA_V_WT     POP_KA     POP_CL BETA_CL_WT   POP_TLAG 
##    7.96620    1.00000    0.35710    0.10026    0.75000    1.00150 
##  POP_FORAL      PPV_V     PPV_KA     PPV_CL   PPV_TLAG  PPV_FORAL 
##    0.98421    0.12539    0.06625    0.12171    0.10000    0.41025 
##  CORR_CL_V    RUV_ADD   RUV_PROP 
##    0.02292    0.00000    0.12309

VPC of model

Before running the VPC with PsN we must update the (initial) values in the MDL Parameter Object MLE estimates from previous step can be used

structuralPar <- getPopulationParameters(NM.FOCEI, what="estimates",block='structural')$MLE
variabilityPar <- getPopulationParameters(NM.FOCEI, what="estimates",block='variability')$MLE

Update the parameter object using the ddmore “updateParObj” function. This function updates an R object of (S4) class “parObj”. The user chooses which block to update, what items within that block, and what to replace those items with. NOTE: that updateParObj can only update attributes which ALREADY EXIST in the MDL Parameter Object for that item. This ensures that valid MDL is preserved.

myParObj <- getParameterObjects(mdlfile)[[1]]
myParObjUpdated <- updateParObj(myParObj,block="STRUCTURAL",
        item=names(structuralPar),
        with=list(value=structuralPar))
myParObjUpdated <- updateParObj(myParObjUpdated,block="VARIABILITY",
        item=names(variabilityPar),
        with=list(value=variabilityPar))

Check that the appropriate initial values have been updated to the MLE values from the previous fit.

print(myParObjUpdated@STRUCTURAL)
## $POP_CL
## $POP_CL$value
## [1] "0.100109"
## 
## $POP_CL$lo
## [1] "0.001"
## 
## 
## $POP_V
## $POP_V$value
## [1] "8.09067"
## 
## $POP_V$lo
## [1] "0.001"
## 
## 
## $POP_KA
## $POP_KA$value
## [1] "0.377193"
## 
## $POP_KA$lo
## [1] "0.001"
## 
## 
## $POP_TLAG
## $POP_TLAG$value
## [1] "1.12074"
## 
## $POP_TLAG$lo
## [1] "0.001"
## 
## 
## $BETA_CL_WT
## $BETA_CL_WT$value
## [1] "0.75"
## 
## $BETA_CL_WT$fix
## [1] "true"
## 
## 
## $BETA_V_WT
## $BETA_V_WT$value
## [1] "1"
## 
## $BETA_V_WT$fix
## [1] "true"
## 
## 
## $POP_FORAL
## $POP_FORAL$value
## [1] "0.999993"
## 
## $POP_FORAL$lo
## [1] "0.001"
print(myParObjUpdated@VARIABILITY)
## $PPV_CL
## $PPV_CL$value
## [1] "0.131938"
## 
## 
## $PPV_V
## $PPV_V$value
## [1] "0.119745"
## 
## 
## $PPV_KA
## $PPV_KA$value
## [1] "0.0350171"
## 
## 
## $PPV_TLAG
## $PPV_TLAG$value
## [1] "0.1"
## 
## $PPV_TLAG$fix
## [1] "true"
## 
## 
## $PPV_FORAL
## $PPV_FORAL$value
## [1] "0.101106"
## 
## 
## $CORR_CL_V
## $CORR_CL_V$value
## [1] "0.0113956"
## 
## 
## $RUV_PROP
## $RUV_PROP$value
## [1] "0.130087"
## 
## $RUV_PROP$lo
## [1] "0"
## 
## 
## $RUV_ADD
## $RUV_ADD$value
## [1] "0.0191873"
## 
## $RUV_ADD$lo
## [1] "1.0E-4"

Assembling the new MOG. Note that we reuse the data and model from the previous run.

myVPCMOG <- createMogObj(dataObj = getDataObjects(mdlfile)[[1]], 
        parObj = myParObjUpdated, 
        mdlObj = getModelObjects(mdlfile)[[1]], 
        taskObj = getTaskPropertiesObjects(mdlfile)[[1]])

We can then write the MOG back out to an .mdl file.

mdlfile.VPC <- paste0(uc,"_VPC.mdl")
writeMogObj(myVPCMOG,mdlfile.VPC)

Similarly as above, ddmore “VPC.PsN” function can be used to run a VPC using PsN as target tool

vpcFiles <- VPC.PsN(mdlfile.VPC,samples=20, seed=12345,
        vpcOptions ="-n_simulation=10",
        subfolder="VPC", plot=TRUE) 
## -- Wed Aug 17 18:43:48 2016
## New
## Submitted
## Job ffa479ab-148a-4c08-89bd-dfc0125a2290 progress:
## Running [ ..... ]
## Importing Results
## Copying the result data back to the local machine for job ID ffa479ab-148a-4c08-89bd-dfc0125a2290...
## From C:\Users\zparra\AppData\Local\Temp\RtmpSKgKTU\DDMORE.job1dec4c0a6c5 to D:/SEE-Prod5_RC4/MDL_IDE/workspace/UseCasesDemo/UseCase4_1/VPC
## Done.
## 
## 
## The following main elements were parsed successfully:
##   RawResults
##   SimulationSimulationBlock
##   SimulationSimulationBlock
## 
## The following MESSAGEs were raised during the job execution:
##  nmoutput2so_version: This SOBlock was created with nmoutput2so version 4.5.27
## 
## Completed
## -- Wed Aug 17 18:45:35 2016

plot of chunk VPC

To replay the visualisation using information from the VPC SO file

pdf(paste0(uc,"_VPC.pdf"))
print(xpose.VPC(vpc.info= file.path("./VPC",vpcFiles@RawResults@DataFiles$PsN_VPC_results$path),
                vpctab= file.path("./VPC",vpcFiles@RawResults@DataFiles$PsN_VPC_vpctab$path),
                main="VPC warfarin"))
dev.off()
## rj.GD 
##     2

Simulation using simulx

The mlxR package has been developed to visualize and explore models that are encoded in MLXTRAN or PharmML. The ddmore function as.PharmML translates an MDL file (extension .mdl) to its PharmML representation. The output file (extension .xml) is saved in the working directory.

myPharmML <- as.PharmML(mdlfile)

Use parameter values from the FOCEI estimation

parValues <- getPopulationParameters(NM.FOCEI,what="estimates")$MLE

Simulate for the typical weight of 70. Recall that logtWT = log(WT/70).

p <- c(parValues,logtWT=0)

Parameter values used in simulation

print(p) 
##     POP_CL      POP_V     POP_KA   POP_TLAG  POP_FORAL   RUV_PROP 
##  0.1001090  8.0906700  0.3771930  1.1207400  0.9999930  0.1300870 
##    RUV_ADD BETA_CL_WT  BETA_V_WT     PPV_CL  CORR_CL_V      PPV_V 
##  0.0191873  0.7500000  1.0000000  0.1319380  0.0113956  0.1197450 
##     PPV_KA   PPV_TLAG  PPV_FORAL     logtWT 
##  0.0350171  0.1000000  0.1011060  0.0000000

Simulate for a dose of 100mg given at time 0 into the CENTRAL (iv administration) and a dose of 150 mg given to the CENTRAL (iv dose). Note that we are using COMPARMTMENTS, which translate to PK macros, and therefore the type option needs to be used

adm1 <- list(type=2, time = 0, amount=100, rate=100) #iv dose
adm2 <- list(type=1, time = 168, amount=150) #oral dose

Simulate PK parameters for individuals

ind <- list(name = c('CL','V'))

Simulate predicted (CONC) and observed concentration values (Y)

f   <- list( name = c('CONC'), time = seq(0,to=300,by=1))
y   <- list( name = c('Y'), time = c(0, 0.5, 1, 4, 8, 12, 24, 36, 48,120,
                168,168.5,170,171,174,180,192,216,240,288))

Simulate 12 subjects

g <- list( size = 12, level = 'individual')

Call simulx

res  <- simulx(model = myPharmML,
        parameter = p,
        group = g,
        treatment = list(adm1,adm2),
        output = list(ind,f,y))

Simulated parameter values for each individual

print(res$parameter)
##    id         CL        V
## 1   1 0.12015483 7.857274
## 2   2 0.10993655 9.953310
## 3   3 0.11213109 7.604861
## 4   4 0.09505271 9.278494
## 5   5 0.09365439 6.492985
## 6   6 0.08721304 8.887331
## 7   7 0.12476215 9.791877
## 8   8 0.10077132 8.550740
## 9   9 0.09184319 9.543819
## 10 10 0.10451931 8.418234
## 11 11 0.11131573 8.664632
## 12 12 0.08643979 9.160808

Plot simulated results

plot(ggplot() + 
                geom_line(data=res$CONC, aes(x=time, y=CONC, colour=id)) +
                geom_point(data=res$Y, aes(x=time, y=Y, colour=id)) +
                xlab("time (h)") + ylab("concentration") )

plot of chunk unnamed-chunk-31

Simulate 1000 subjects - with simulx this is a QUICK process!

g <- list( size = 1000, level = 'individual')

Call simulx

res.1000  <- simulx(model = myPharmML,
        parameter = p,
        group = g,
        treatment = list(adm1,adm2),
        output = list(ind,f,y))

Plot prediction intervals with prctilemlx. band defines the percentile bands displayed:

print(prctilemlx(res.1000$Y,band=list(number=9, level=90)))

plot of chunk unnamed-chunk-33

Table of the same information

print(prctilemlx(res.1000$Y,band=list(number=10, level=100), plot=FALSE)$y)
##     time          0%        10%         20%          30%          40%
## 1    0.0 -0.04250214 -0.0214894 -0.01205953 -0.007334241 -0.003588708
## 2    0.5  3.45658478  4.5928096  4.96053979  5.375915594  5.617766642
## 3    1.0  8.13273464 10.1160323 10.71280121 11.118709044 11.608872919
## 4    4.0  7.76795223  9.3814666 10.25757792 10.619837653 11.099371981
## 5    8.0  7.58396685  9.1939082  9.76168375 10.301531371 10.859190923
## 6   12.0  6.41657444  8.3821669  9.14008881  9.516837941 10.040150210
## 7   24.0  6.56708633  7.5726987  7.93899750  8.473413943  8.783496184
## 8   36.0  5.06547333  6.3247464  6.95721820  7.343613967  7.560217130
## 9   48.0  4.06691053  5.0977205  5.79851816  6.103610306  6.314033604
## 10 120.0  1.41476758  2.0538022  2.21436078  2.389640975  2.656988092
## 11 168.0  0.79806050  0.9534254  1.23424933  1.371211220  1.460617041
## 12 168.5  0.76305786  0.9925449  1.18908523  1.326614754  1.452759845
## 13 170.0  4.47187982  5.1638308  5.51804862  5.848663634  6.229930178
## 14 171.0  6.60132336  8.6620018  9.20458206  9.844003692 10.204282703
## 15 174.0  9.12585203 13.1524954 14.21175593 15.054218435 15.637885742
## 16 180.0 11.10045309 14.4411164 15.08506737 16.166610591 16.885898936
## 17 192.0  9.03859184 12.3639475 13.75531721 14.498626976 14.918112816
## 18 216.0  7.32080496  9.4262119  9.93510227 10.483802212 11.010911050
## 19 240.0  4.57663723  6.8208800  7.11578286  7.434450762  7.961191078
## 20 288.0  2.50389135  3.4898613  3.95353595  4.224350545  4.464408512
##             50%          50%          60%          70%         80%
## 1  -0.001383821 -0.001383821  0.003274618  0.006086306  0.01214598
## 2   5.875382617  5.875382617  6.113136350  6.572341863  6.95496915
## 3  12.222760808 12.222760808 12.703412299 13.079425300 14.10999083
## 4  11.430317847 11.430317847 11.688034221 12.339419305 13.09266285
## 5  11.243483470 11.243483470 11.578293795 12.120804079 12.46764701
## 6  10.598746060 10.598746060 11.037085213 11.480117689 11.83416698
## 7   9.056183297  9.056183297  9.306928797  9.581069929 10.14503300
## 8   7.812607636  7.812607636  8.162125170  8.633432235  9.02913318
## 9   6.687683638  6.687683638  6.985837460  7.286619361  7.65221006
## 10  2.886398188  2.886398188  2.991237750  3.185885315  3.41723779
## 11  1.624727805  1.624727805  1.698117627  1.833057810  1.96672351
## 12  1.538191009  1.538191009  1.625924068  1.731828740  1.92382845
## 13  6.508340387  6.508340387  6.729499398  7.143008812  7.68539598
## 14 10.636268514 10.636268514 10.964340502 11.286458961 12.06295321
## 15 16.211919320 16.211919320 16.950453799 17.685471515 18.22459712
## 16 17.715475460 17.715475460 18.131207903 18.869008072 20.23348295
## 17 15.218370877 15.218370877 15.654460392 16.218587052 17.35070931
## 18 11.399055974 11.399055974 11.868962995 12.537849592 13.05348113
## 19  8.466998008  8.466998008  8.789664745  9.016811333  9.88939064
## 20  4.678355503  4.678355503  4.944535859  5.187530239  5.54355470
##            90%       100%
## 1   0.02709967  0.0362757
## 2   7.53863254  9.8121637
## 3  15.02200955 17.4222949
## 4  14.14538728 17.1762858
## 5  13.11002937 14.9513483
## 6  12.63641439 14.6670366
## 7  10.86130713 12.4234564
## 8   9.42862106 11.8810324
## 9   8.09348330  9.2101867
## 10  3.73882855  4.6292115
## 11  2.36743710  3.1396415
## 12  2.16646542  2.7571297
## 13  8.21016126  9.5130197
## 14 12.80749907 14.4194950
## 15 19.37675160 22.3810395
## 16 21.38023288 25.1724714
## 17 18.03688079 23.4713964
## 18 14.18492279 16.7574227
## 19 10.55146404 13.5408111
## 20  6.11301564  8.1673443