examples/design-method-simulateDualResponsesSamplesDesign.R

In Roche/crmPack: Object-Oriented Implementation of CRM Designs

# nolint start

## Simulate dose-escalation procedure based on DLE and efficacy responses where DLE
## and efficacy samples are used
data <- DataDual(doseGrid = seq(25, 300, 25), placebo = FALSE)
## First for the DLE model
## The DLE model must be of 'ModelTox' (e.g 'LogisticIndepBeta') class
DLEmodel <- LogisticIndepBeta(
  binDLE = c(1.05, 1.8),
  DLEweights = c(3, 3),
  DLEdose = c(25, 300),
  data = data
)

## The efficacy model of 'ModelEff' (e.g 'Effloglog') class
Effmodel <- Effloglog(
  eff = c(1.223, 2.513), eff_dose = c(25, 300),
  nu = c(a = 1, b = 0.025), data = data
)


## The escalation rule using the 'NextBestMaxGainSamples' class
mynextbest <- NextBestMaxGainSamples(
  prob_target_drt = 0.35,
  prob_target_eot = 0.3,
  derive = function(samples) {
    as.numeric(quantile(samples, prob = 0.3))
  },
  mg_derive = function(mg_samples) {
    as.numeric(quantile(mg_samples, prob = 0.5))
  }
)

## The increments (see Increments class examples)
## 200% allowable increase for dose below 300 and 200% increase for dose above 300
myIncrements <- IncrementsRelative(
  intervals = c(25, 300),
  increments = c(2, 2)
)
## cohort size of 3
mySize <- CohortSizeConst(size = 3)
## Stop only when 10 subjects are treated (only for illustration such a low
## sample size)
myStopping <- StoppingMinPatients(nPatients = 10)
## Now specified the design with all the above information and starting with
## a dose of 25


## Specified the design
design <- DualResponsesSamplesDesign(
  nextBest = mynextbest,
  cohort_size = mySize,
  startingDose = 25,
  model = DLEmodel,
  eff_model = Effmodel,
  data = data,
  stopping = myStopping,
  increments = myIncrements
)
## specified the true DLE and efficacy curve
myTruthDLE <- probFunction(DLEmodel, phi1 = -53.66584, phi2 = 10.50499)
myTruthEff <- efficacyFunction(Effmodel, theta1 = -4.818429, theta2 = 3.653058)

## The true gain curve can also be seen
myTruthGain <- function(dose) {
  return((myTruthEff(dose)) / (1 + (myTruthDLE(dose) / (1 - myTruthDLE(dose)))))
}

## simulate the trial for 10 times involving samples
## for illustration purpose we use 10 burn-ins to generate 50 samples
options <- McmcOptions(burnin = 10, step = 1, samples = 50)
## For illustration purpose only 1 simulations are produced (nsim=1).
mySim <- simulate(design,
  args = NULL,
  trueDLE = myTruthDLE,
  trueEff = myTruthEff,
  trueNu = 1 / 0.025,
  nsim = 1,
  mcmcOptions = options,
  seed = 819,
  parallel = FALSE
)



## Simulate dose-escalation procedure based on DLE and efficacy responses where DLE
## and efficacy samples are used
## when the efficacy model is of 'EffFlexi' class
Effmodel <- EffFlexi(
  eff = c(1.223, 2.513), eff_dose = c(25, 300),
  sigma2W = c(a = 0.1, b = 0.1), sigma2betaW = c(a = 20, b = 50), rw1 = FALSE, data = data
)


## Specified the design
design <- DualResponsesSamplesDesign(
  nextBest = mynextbest,
  cohort_size = mySize,
  startingDose = 25,
  model = DLEmodel,
  eff_model = Effmodel,
  data = data,
  stopping = myStopping,
  increments = myIncrements
)
## specified the true DLE curve and the true expected efficacy values at all dose levels
myTruthDLE <- probFunction(DLEmodel, phi1 = -53.66584, phi2 = 10.50499)

myTruthEff <- c(
  -0.5478867, 0.1645417, 0.5248031, 0.7604467,
  0.9333009, 1.0687031, 1.1793942, 1.2726408,
  1.3529598, 1.4233411, 1.4858613, 1.5420182
)
## The true gain curve can also be seen
d1 <- data@doseGrid
myTruthGain <- (myTruthEff) / (1 + (myTruthDLE(d1) / (1 - myTruthDLE(d1))))


mySim <- simulate(
  object = design,
  args = NULL,
  trueDLE = myTruthDLE,
  trueEff = myTruthEff,
  trueSigma2 = 0.025,
  trueSigma2betaW = 1,
  mcmcOptions = options,
  nsim = 1,
  seed = 819,
  parallel = FALSE
)

# nolint end