learn_MHN: Learn a Mutual Hazard Network from a set of mutation trees

In cbg-ethz/TreeMHN: Joint inference of repeated evolutionary trajectories and patterns of clonal exclusivity from single-cell mutation trees

Learn a Mutual Hazard Network from a set of mutation trees

Description

This function learns a Mutual Hazard Network from a set of mutation trees in the format of a TreeMHN object.

Usage

learn_MHN(
  tree_obj,
  gamma = 0.5,
  lambda_s = 1,
  Theta_init = NULL,
  M = 100,
  iterations = 500,
  to_mask = integer(0),
  use_EM = FALSE,
  verbose = FALSE,
  MC_threshold = 500,
  increment_M = 0,
  increment_M_bound = 500,
  return_Theta_only = TRUE
)

Arguments

tree_obj	A TreeMHN object
gamma	Penalization parameter in the objective function (Default: 0.5).
lambda_s	Sampling rate (Default: 1).
Theta_init	Initial value of the MHN provided to the optimization procedure (Default: NULL).
M	Number of Monte Carlo samples to be drawn (Default: 100).
iterations	Number of iterations for the EM/MCEM algorithm (Default: 500).
to_mask	An integer vector of indices by column, which is used to mask the off-diagonal entries of an MHN (Default: an empty vector).
use_EM	A boolean value to determine whether the EM/MCEM algorithm is used (Default: FALSE).
verbose	A boolean value to determine whether optimization steps are printed (Default: FALSE).
MC_threshold	A threshold on the maximum number of subtrees of a given tree, above which Monte Carlo sampling will be used (Default: 500).
increment_M	The step size to increment the number of Monte Carlo samples (Default: 0).
increment_M_bound	The upper bound on the number of Monte Carlo samples (Default: 500).
return_Theta_only	A boolean value to determine whether the function returns only the estimated Theta or the TreeMHN object containing the estimated Theta and other parameters (Default: TRUE).

Value

A Mutual Hazard Network Theta

Author(s)

Xiang Ge Luo

cbg-ethz/TreeMHN documentation built on Jan. 29, 2024, 1:29 p.m.