![[Experimental]](figures/lifecycle-experimental.svg)
TwoDrugsCombo is the class for a two-drug combination regression model with
fixed priors for the two single-agent dose-toxicity models and an additional
interaction parameter.
Usage
TwoDrugsCombo(single_models, gamma = 0, tau = 1, log_normal_eta = FALSE)
.DefaultTwoDrugsCombo()Arguments
- single_models
(
list) named list of length 2 with compatible single-agentGeneralModelobjects, one per drug.- gamma
(
number) prior mean parameter for the interaction term.- tau
(
number) prior precision parameter for the interaction term.- log_normal_eta
(
flag) should the interaction term use a log-normal prior?
Details
Let \(p(x_1, x_2)\) be the probability of DLT at the dose
combination \((x_1, x_2)\). The model combines two single-agent models
with an interaction term:
$$\textrm{odds}(p(x_1, x_2)) = \textrm{odds}(p_0(x_1, x_2)) *
\exp\left(\eta * I(x_1, x_2)\right),$$
where \(p_0(x_1, x_2) = 1 - (1 - p_1(x_1))(1 - p_2(x_2))\) and each
single-agent probability follows a model \(p_j(x_j)\). The normalized dose
\(\tilde{x}_j\) is extracted from the single-agent model's dose covariate,
e.g. \(x_j / x_j^{*}\), \(x_j - x_j^{*}\), or \(x_j\).
The interaction parameter \(\eta\) has either a normal prior or, if
log_normal_eta = TRUE, a log-normal prior.
Slots
single_models(
list) named list of length 2 containing single-agentGeneralModelobjects, one per drug. Each model must usenObs,y, andxas data inputs and contain a Bernoulli likelihood foryin itsdatamodel.ref_dose(
numeric) optional reference doses extracted fromsingle_models, if provided.drug_names(
character) the names of the two drugs.gamma(
numeric) prior mean parameter for the interaction term.tau(
numeric) prior precision parameter for the interaction term.log_normal_eta(
flag) should the interaction term use a log-normal prior?
Examples
my_model <- TwoDrugsCombo(
single_models = list(
drug1 = LogisticLogNormal(
mean = c(-0.85, 1),
cov = matrix(c(1, -0.5, -0.5, 1), nrow = 2),
ref_dose = 10
),
drug2 = LogisticLogNormal(
mean = c(-0.7, 0.8),
cov = matrix(c(1.1, -0.3, -0.3, 0.9), nrow = 2),
ref_dose = 20
)
),
gamma = 0,
tau = 1
)
my_model
#> An object of class "TwoDrugsCombo"
#> Slot "single_models":
#> $drug1
#> An object of class "LogisticLogNormal"
#> Slot "params":
#> An object of class "ModelParamsNormal"
#> Slot "mean":
#> [1] -0.85 1.00
#>
#> Slot "cov":
#> [,1] [,2]
#> [1,] 1.0 -0.5
#> [2,] -0.5 1.0
#>
#> Slot "prec":
#> [,1] [,2]
#> [1,] 1.3333333 0.6666667
#> [2,] 0.6666667 1.3333333
#>
#>
#> Slot "ref_dose":
#> An object of class "positive_number"
#> [1] 10
#>
#> Slot "datamodel":
#> function() {
#> for (i in 1:nObs) {
#> logit(p[i]) <- alpha0 + alpha1 * log(x[i] / ref_dose)
#> y[i] ~ dbern(p[i])
#> }
#> }
#> <bytecode: 0x560dba92b748>
#> <environment: 0x560dc20f6a70>
#>
#> Slot "priormodel":
#> function() {
#> theta ~ dmnorm(mean, prec)
#> alpha0 <- theta[1]
#> alpha1 <- exp(theta[2])
#> }
#> <bytecode: 0x560dbb4b0988>
#> <environment: 0x560dc20f6840>
#>
#> Slot "modelspecs":
#> function(from_prior) {
#> ms <- list(mean = params@mean, prec = params@prec)
#> if (!from_prior) {
#> ms$ref_dose <- ref_dose
#> }
#> ms
#> }
#> <bytecode: 0x560dbb9b8710>
#> <environment: 0x560dc20f6840>
#>
#> Slot "init":
#> function() {
#> list(theta = c(0, 1))
#> }
#> <bytecode: 0x560dbba3ab18>
#> <environment: 0x560dc20f6840>
#>
#> Slot "datanames":
#> [1] "nObs" "y" "x"
#>
#> Slot "datanames_prior":
#> character(0)
#>
#> Slot "sample":
#> [1] "alpha0" "alpha1"
#>
#>
#> $drug2
#> An object of class "LogisticLogNormal"
#> Slot "params":
#> An object of class "ModelParamsNormal"
#> Slot "mean":
#> [1] -0.7 0.8
#>
#> Slot "cov":
#> [,1] [,2]
#> [1,] 1.1 -0.3
#> [2,] -0.3 0.9
#>
#> Slot "prec":
#> [,1] [,2]
#> [1,] 1.0000000 0.3333333
#> [2,] 0.3333333 1.2222222
#>
#>
#> Slot "ref_dose":
#> An object of class "positive_number"
#> [1] 20
#>
#> Slot "datamodel":
#> function() {
#> for (i in 1:nObs) {
#> logit(p[i]) <- alpha0 + alpha1 * log(x[i] / ref_dose)
#> y[i] ~ dbern(p[i])
#> }
#> }
#> <bytecode: 0x560dba92b748>
#> <environment: 0x560dc4a07ba8>
#>
#> Slot "priormodel":
#> function() {
#> theta ~ dmnorm(mean, prec)
#> alpha0 <- theta[1]
#> alpha1 <- exp(theta[2])
#> }
#> <bytecode: 0x560dbb4b0988>
#> <environment: 0x560dc4a07978>
#>
#> Slot "modelspecs":
#> function(from_prior) {
#> ms <- list(mean = params@mean, prec = params@prec)
#> if (!from_prior) {
#> ms$ref_dose <- ref_dose
#> }
#> ms
#> }
#> <bytecode: 0x560dbb9b8710>
#> <environment: 0x560dc4a07978>
#>
#> Slot "init":
#> function() {
#> list(theta = c(0, 1))
#> }
#> <bytecode: 0x560dbba3ab18>
#> <environment: 0x560dc4a07978>
#>
#> Slot "datanames":
#> [1] "nObs" "y" "x"
#>
#> Slot "datanames_prior":
#> character(0)
#>
#> Slot "sample":
#> [1] "alpha0" "alpha1"
#>
#>
#>
#> Slot "ref_dose":
#> drug1 drug2
#> 10 20
#>
#> Slot "drug_names":
#> [1] "drug1" "drug2"
#>
#> Slot "gamma":
#> [1] 0
#>
#> Slot "tau":
#> [1] 1
#>
#> Slot "log_normal_eta":
#> [1] FALSE
#>
#> Slot "datamodel":
#> function ()
#> {
#> for (i in 1:nObs) {
#> x_drug1[i] <- x[i, 1L]
#> }
#> for (i in 1:nObs) {
#> logit(p_drug1[i]) <- alpha0_drug1 + alpha1_drug1 * log(x_drug1[i]/ref_dose_drug1)
#> p_single[i, 1L] <- p_drug1[i]
#> }
#> for (i in 1:nObs) {
#> x_drug2[i] <- x[i, 2L]
#> }
#> for (i in 1:nObs) {
#> logit(p_drug2[i]) <- alpha0_drug2 + alpha1_drug2 * log(x_drug2[i]/ref_dose_drug2)
#> p_single[i, 2L] <- p_drug2[i]
#> }
#> for (i in 1:nObs) {
#> combo_interaction[i] <- x_drug1[i]/ref_dose_drug1 * (x_drug2[i]/ref_dose_drug2)
#> }
#> for (i in 1:nObs) {
#> p0[i] <- p_single[i, 1] + p_single[i, 2] - p_single[i,
#> 1] * p_single[i, 2]
#> logit(p[i]) <- log(p0[i]/(1 - p0[i])) + eta * combo_interaction[i]
#> y[i] ~ dbern(p[i])
#> }
#> }
#> <environment: 0x560dc66502e8>
#>
#> Slot "priormodel":
#> function ()
#> {
#> theta_drug1 ~ dmnorm(mean_drug1, prec_drug1)
#> alpha0_drug1 <- theta_drug1[1]
#> alpha1_drug1 <- exp(theta_drug1[2])
#> theta_drug2 ~ dmnorm(mean_drug2, prec_drug2)
#> alpha0_drug2 <- theta_drug2[1]
#> alpha1_drug2 <- exp(theta_drug2[2])
#> alpha0[1L] <- alpha0_drug1
#> alpha0[2L] <- alpha0_drug2
#> alpha1[1L] <- alpha1_drug1
#> alpha1[2L] <- alpha1_drug2
#> eta ~ dnorm(eta_gamma, eta_tau)
#> }
#> <environment: 0x560dc20f6840>
#>
#> Slot "modelspecs":
#> function(from_prior) {
#> specs_name <- if (from_prior) "prior_specs" else "full_specs"
#> ms <- c(
#> unlist(lapply(single_model_parts, "[[", specs_name), recursive = FALSE),
#> list(eta_gamma = gamma, eta_tau = tau)
#> )
#> ms
#> }
#> <bytecode: 0x560db41314d0>
#> <environment: 0x560dc20f7bf0>
#>
#> Slot "init":
#> function() {
#> c(
#> unlist(lapply(single_model_parts, "[[", "inits"), recursive = FALSE),
#> list(eta = gamma)
#> )
#> }
#> <bytecode: 0x560db3a344a8>
#> <environment: 0x560dc20f7bf0>
#>
#> Slot "datanames":
#> [1] "nObs" "y" "x"
#>
#> Slot "datanames_prior":
#> character(0)
#>
#> Slot "sample":
#> [1] "alpha0" "alpha1" "eta"
#>