variance_target: Second adaption phase variance tuning

In anna-pacinkova/intomics_package: Integrative analysis of multi-omics data to infer regulatory networks

Second adaption phase variance tuning

Description

variance_target This phase identifies the proposal distribution that has a similar covariance structure with the target distribution. This is part of second_adapt_phase.

Usage

variance_target(
  transient.phase_net,
  constant,
  fin,
  B_prior_mat,
  omics,
  parent_set_combinations,
  BGe_score_all_configs_node,
  layers_def,
  prob_mbr,
  annot
)

Arguments

transient.phase_net	list output of the variance_target or transient.phase function.
constant	numeric vector used to multiply the beta_sd to determine the variance of the distribution of the hyperparameter beta.
fin	numeric vector iteration to stop.
B_prior_mat	a biological prior matrix.
omics	named list containing the gene expression (possibly copy number variation and methylation data). Each component of the list is a matrix with samples in rows and features in columns.
parent_set_combinations	list of all possible parent set configuration for all nodes available.
BGe_score_all_configs_node	list of nodes BGe score for all possible parent set configurations.
layers_def	data.frame containing the modality ID, corresponding layer in BN and maximal number of parents from given layer to GE nodes.
prob_mbr	numeric vector probability of the MBR step.
annot	named list containing the associated methylation probes of given gene.

Value

Large List of 3 elements: second adaptive phase result with possible MCMC mixing; acceptance rate of hyperparameter beta; SD of hyperparameter beta

Examples

data(list=c("PK", "TFtarg_mat", "annot", "layers_def", "omics"),
    package="IntOMICS")
OMICS_mod_res <- OMICS_module(omics = omics, PK = PK, annot = annot, 
    layers_def = layers_def, TFtargs = TFtarg_mat, r_squared_thres = 0.3,
    lm_METH = TRUE)
first.adapt.phase_net <- first_adapt_phase(omics = OMICS_mod_res$omics, 
    B_prior_mat = OMICS_mod_res$B_prior_mat, 
    energy_all_configs_node = OMICS_mod_res$pf_UB_BGe_pre$energy_all_configs_node,
    len = 5, layers_def = OMICS_mod_res$layers_def, prob_mbr = 0.07,
    BGe_score_all_configs_node = OMICS_mod_res$pf_UB_BGe_pre$BGe_score_all_configs_node, 
    parent_set_combinations = OMICS_mod_res$pf_UB_BGe_pre$parents_set_combinations, 
    annot = OMICS_mod_res$annot)
transient.phase_net <- transient_phase(annot = OMICS_mod_res$annot, 
    first.adapt.phase_net = first.adapt.phase_net, 
    omics = OMICS_mod_res$omics, prob_mbr = 0.07, 
    B_prior_mat = OMICS_mod_res$B_prior_mat, 
    layers_def = OMICS_mod_res$layers_def, 
    energy_all_configs_node = OMICS_mod_res$pf_UB_BGe_pre$energy_all_configs_node,
    BGe_score_all_configs_node = OMICS_mod_res$pf_UB_BGe_pre$BGe_score_all_configs_node, 
    parent_set_combinations = OMICS_mod_res$pf_UB_BGe_pre$parents_set_combinations) 
variance_target(transient.phase_net = transient.phase_net, 
    constant = 1.586667, fin = 200, B_prior_mat = OMICS_mod_res$B_prior_mat,
    parent_set_combinations = OMICS_mod_res$pf_UB_BGe_pre$parents_set_combinations, 
    BGe_score_all_configs_node = OMICS_mod_res$pf_UB_BGe_pre$BGe_score_all_configs_node, 
    layers_def = OMICS_mod_res$layers_def, omics = OMICS_mod_res$omics, 
    prob_mbr = 0.07, annot = OMICS_mod_res$annot)

anna-pacinkova/intomics_package documentation built on Aug. 13, 2022, 11:38 a.m.