Extracts regression coefficients from TOP posterior samples

extract_TOP_coef_samples(
  TOP_samples,
  tf_cell_combos,
  assembled_training_data,
  tf_name,
  cell_type,
  n_bins = 5,
  level = c("bottom", "middle", "top")
)

Arguments

TOP_samples: TOP samples combined from all partitions using the combine_TOP_samples() function.
tf_cell_combos: A table with the indices and names of TF and cell type combinations from the assembled training data. If missing, will create the table using assembled_training_data.
assembled_training_data: Assembled training data as in the assemble_training_data() function.
tf_name: TF name.
cell_type: Cell type.
n_bins: Number of DNase or ATAC bins in TOP model (default = 5).
level: The level in the TOP model (options: ‘bottom’, ‘middle’, or ‘top’). ‘bottom’ level: TF- and cell-type- specific; ‘middle’ level: TF-specific, cell-type generic; ‘top’ level: TF-generic

Value

A data frame of posterior samples for TOP's regression coefficients.

Examples

if (FALSE) {
TOP_coef_samples <- extract_TOP_coef_samples(TOP_samples,
                                             assembled_training_data = assembled_training_data)
# or
tf_cell_combos <- extract_tf_cell_combos(assembled_training_data)
TOP_coef_samples <- extract_TOP_coef_samples(TOP_samples,
                                             tf_cell_combos = tf_cell_combos)
}