Execution of a Single Chain in Metropolis-Hastings for Cancer Risk Estimation

Performs a single chain execution in the Metropolis-Hastings algorithm for Bayesian inference, specifically tailored for cancer risk estimation. This function can handle both sex-specific and non-sex-specific scenarios.

Usage

mhChain(
  seed,
  n_iter,
  burn_in,
  chain_id,
  ncores,
  data,
  twins,
  max_age,
  baseline_data,
  prior_distributions,
  prev,
  median_max,
  BaselineNC,
  var,
  age_imputation,
  imp_interval,
  remove_proband,
  sex_specific
)

Arguments

seed: Integer, the seed for the random number generator to ensure reproducibility.
n_iter: Integer, the number of iterations to perform in the Metropolis-Hastings algorithm.
burn_in: Integer, the number of initial iterations to discard (burn-in period).
chain_id: Integer, the identifier for the chain being executed.
ncores: Integer, the number of cores to use for parallel computation.
data: Data frame, containing family and genetic information used in the analysis.
twins: Information on monozygous twins or triplets in the pedigrees.
max_age: Integer, the maximum age considered in the analysis.
baseline_data: Numeric matrix or vector, containing baseline risk estimates for different ages and sexes.
prior_distributions: List, containing prior distributions for the parameters being estimated.
prev: Numeric, the prevalence of the risk allele in the population.
median_max: Logical, indicates if the maximum median age should be used for the Weibull distribution.
BaselineNC: Logical, indicates if non-carrier penetrance should be based on SEER data.
var: Numeric, the variance for the proposal distribution in the Metropolis-Hastings algorithm.
age_imputation: Logical, indicates if age imputation should be performed.
imp_interval: Integer, the interval at which age imputation should be performed when age_imputation = TRUE.
remove_proband: Logical, indicates if the proband should be removed from the analysis.
sex_specific: Logical, indicates if the analysis should differentiate by sex.

Value

A list containing samples, log likelihoods, log-acceptance ratio, and rejection rate for each iteration.