This is my modified version of the original OncoBayes2 package.
The most significant difference is the support of providing correlation
parameter when sampling the component via the newly added parameter
prior_EX_corr_mu_comp in blrm_exnex().
In the original package,
prior_EX_mu_sd_comp(Note, this parameter provides the
standard variance).
Now, the off-diagonal part of prior_EX_corr_mu_comp.
-
This is just a proof of concept during my replication of Table 14-7 in this trial protocol. The original OncoBayes2 misses the functionality to accept user defined correlation, which is also reported on the internet
-
Currently this package is forked and modified from CRAN’s github version of OncoBayes2. After installation it will overwrite your existing OncoBayes2.
-
This is currently WIP. Many aspect, like the
prior_summarymethod should be updated accordingly to reflect the correlation information. Also many document template is missing hence the manual cannot be auto-updated properly.
You can install the development version of my modified OncoBayes2 like so:
remotes::install_github("fenguoerbian/MyOncoBayes2")This is a basic example which reproduce the results in Table 14-7 in this trial protocol.
library(OncoBayes2)
#> This is OncoBayes2 version 0.8.123456 (released 2023-07-21, git-sha e5a8128)set.seed(123)
.user_mc_options <- options(OncoBayes2.MC.warmup=500, OncoBayes2.MC.iter=20000, OncoBayes2.MC.chains=10,
OncoBayes2.MC.save_warmup=FALSE, mc.cores = 10)SA_trial_setup <- blrm_trial(
data = tibble::tibble(group_id = as.factor("All"),
drug_A = 50,
num_patients = NA,
num_toxicities = NA),
# dose-toxicity data available at design stage of trial.
dose_info = tibble::tibble(group_id = as.factor("All"),
drug_A = c(50,75,150,300,450,600,800,1000),
dose_id = c(1, 2, 3, 4, 5, 6, 7, 8), stratum_id = "all"),
# specification of the dose levels as planned for the ongoing trial arms.
drug_info = tibble::tibble(drug_name = "drug_A",
dose_ref = 300,
dose_unit = "ug/kg"
,
reference_p_dlt = exp(-3.068) / (1 + exp(-3.068))
)
# specification of drugs used in trial arms
,simplified_prior = FALSE
,EXNEX_comp=TRUE
,EXNEX_inter=FALSE
,interval_prob = c(0,0.16,0.33,1)
,interval_max_mass = c(prob_underdose = 1 # The prob of under-dose is allowed maximum to 1.
, prob_target = 1 # The prob of target-dose is allowed maximum to 1.
, prob_overdose = 0.25) # The prob of over-dose is allowed maximum to 0.28.
)
#> Please configure blrm_exnex using the update() function.
dims <- summary(SA_trial_setup,"dimensionality")
num_comp <- dims$num_componentsdims <- summary(SA_trial_setup,"dimensionality")
num_comp <- dims$num_components
SA_trial_start <- update(
SA_trial_setup,
# component1 MAP prior
# check the usage in single agent example documentation.
prior_EX_mu_mean_comp = matrix(
c(-3.068, # mean of intercept
0.564), # mean of log-slope
nrow = num_comp,
ncol = 2
),
prior_EX_mu_sd_comp = matrix(
c(2.706, # sd of intercept
0.728), # sd of log-slope
nrow = num_comp,
ncol = 2
),
prior_EX_corr_mu_comp = -0.817, # HERE! supply the CORRELATION!
prior_EX_tau_mean_comp = matrix(
c(0, 0),
nrow = num_comp,
ncol = 2
),
prior_EX_tau_sd_comp = matrix(
c(1, 1),
nrow = num_comp,
ncol = 2
),
prior_EX_prob_comp = matrix(1, nrow = 1, ncol = 1),
prior_tau_dist = 0, # single-agent, without historical data, set this to 0.
# see single-agent example in this package for more details
prior_PD = FALSE
)NOTE: prior_summary needs to be updated to reflect user supplied
correlation
prior_summary(SA_trial_start)
#> Bayesian Logistic Regression Model with EXchangeability-NonEXchangeability
#>
#> Mixture configuration
#> ---------------------
#> EXNEX components : 0
#> component
#> I(log(drug_A/300))
#> 0
#>
#> Prior probability for exchangeability per group
#> component
#> group I(log(drug_A/300))
#> All 1
#>
#> EXchangable hyperparameter priors
#> ---------------------------------
#> Component parameters
#> Mean mu_log_beta
#> coefficient intercept log_slope
#> prior mean sd mean sd
#> component
#> I(log(drug_A/300)) -3.07 2.71 0.56 0.73
#>
#> Heterogeneity tau_log_beta (known)
#> coefficient intercept log_slope
#> stratum all all
#> prior mean sd mean sd
#> component
#> I(log(drug_A/300)) 0 1 0 1
#>
#> Correlation LKJ
#> component
#> I(log(drug_A/300))
#> 1
#>
#> Model has no interaction parameters.
#>
#> NonEXchangable priors
#> ---------------------
#> Component parameters
#> Mean mu_log_beta
#> coefficient intercept log_slope
#> prior mean sd mean sd
#> component
#> I(log(drug_A/300)) -3.07 2.71 0.56 0.73
#>
#> Model has no interaction parameters.summary(SA_trial_start, "dose_prediction")
#> # A tibble: 8 × 13
#> group_id drug_A dose_id stratum_id mean sd `2.5%` `50%` `97.5%` prob_underdose prob_target prob_overdose ewoc_ok
#> <fct> <dbl> <dbl> <fct> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <lgl>
#> 1 All 50 1 all 0.0776 0.190 8.75e-10 0.00183 0.786 0.867 0.0484 0.0843 TRUE
#> 2 All 75 2 all 0.0905 0.202 1.61e- 8 0.00372 0.819 0.845 0.0558 0.0993 TRUE
#> 3 All 150 3 all 0.121 0.226 2.23e- 6 0.0128 0.868 0.791 0.0747 0.134 TRUE
#> 4 All 300 4 all 0.172 0.254 2.27e- 4 0.0447 0.907 0.698 0.110 0.192 TRUE
#> 5 All 450 5 all 0.223 0.269 2.02e- 3 0.0961 0.926 0.602 0.146 0.253 FALSE
#> 6 All 600 6 all 0.282 0.279 5.91e- 3 0.172 0.939 0.483 0.185 0.332 FALSE
#> 7 All 800 7 all 0.366 0.291 1.37e- 2 0.289 0.959 0.335 0.207 0.457 FALSE
#> 8 All 1000 8 all 0.441 0.298 2.37e- 2 0.400 0.975 0.235 0.197 0.569 FALSEsummary(SA_trial_start, "ewoc_check")
#> # A tibble: 8 × 9
#> group_id drug_A dose_id stratum_id prob_overdose_est prob_overdose_stat prob_overdose_mcse prob_overdose_ess prob_overdose_rhat
#> <fct> <dbl> <dbl> <fct> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 All 50 1 all 0.0336 -609. 0.000487 116842. 1.00
#> 2 All 75 2 all 0.0525 -379. 0.000732 117018. 1.00
#> 3 All 150 3 all 0.110 -172. 0.00128 119441. 1.00
#> 4 All 300 4 all 0.225 -57.4 0.00184 123470. 1.00
#> 5 All 450 5 all 0.335 2.33 0.00198 127508. 1.00
#> 6 All 600 6 all 0.447 58.5 0.00200 133847. 1.00
#> 7 All 800 7 all 0.589 149. 0.00174 138891. 1.00
#> 8 All 1000 8 all 0.696 258. 0.00142 144318. 1.00