TwoDrugsCombo-class: `TwoDrugsCombo`

Description

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-agent GeneralModel objects, 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?

Slots

single_models: (list) named list of length 2 containing single-agent GeneralModel objects, one per drug. Each model must use nObs, y, and x as data inputs and contain a Bernoulli likelihood for y in its datamodel.

ref_dose

(numeric) optional reference doses extracted from single_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?

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.

Examples

Run this code

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