Relearn Trajectory Model — relearn_traj

This function relearns a trajectory model using the glmnet package, without re-inferring the motif models.

Usage

relearn_traj_model(
  traj_model,
  new_energies = FALSE,
  new_logist = FALSE,
  lambda = NULL,
  use_additional_features = TRUE,
  use_motifs = TRUE,
  verbose = FALSE,
  rescale_pred = TRUE,
  relearn_model = TRUE,
  family = "binomial",
  new_interactions = FALSE,
  max_n_interactions = NULL,
  use_cv = FALSE,
  nfolds = 10,
  interaction_scale_factor = 1,
  logist_interactions = FALSE,
  logist_dinucs = FALSE,
  new_score = NULL
)

Arguments

traj_model: The trajectory model object.
new_energies: If TRUE - recreate the energies. Default is FALSE.
new_logist: If TRUE - recreate the logistic features. Default is FALSE. If new_energies is TRUE, this is automatically set to TRUE.
lambda: The lambda value to use for relearning. If NULL, the lambda value from the trajectory model is used.
use_additional_features: A logical value indicating whether to use additional features. Default is TRUE.
use_motifs: A logical value indicating whether to use motif models. Default is TRUE.
verbose: Logical indicating whether to display additional information.
rescale_pred: Logical indicating whether to rescale the predicted values. Default is TRUE.
relearn_model: Logical indicating whether to relearn the model. Default is TRUE.
family: The family to use for the glmnet model. Either "binomial" (default) or "gaussian" for linear regression.
new_interactions: Logical indicating whether to add new interactions. Default is FALSE.
max_n_interactions: The maximum number of interactions to add. Default is NULL.
use_cv: Logical indicating whether to use cross-validation for lambda selection. Default is FALSE.
nfolds: Number of folds for cross-validation. Default is 10.
interaction_scale_factor: The factor to scale the interactions by. Default is 1 (no scaling).
logist_interactions: Logical indicating whether to transform interactions to logistic functions. Default is FALSE.
logist_dinucs: Logical indicating whether to transform dinucleotides to logistic functions. Default is FALSE.
new_score: A vector of new scores to use for the model. If NULL, the diff_score is used. Default is NULL. If the vector is named, the names should be the same as the peak names in the trajectory model.

Value

The updated trajectory model object.