Learn R Programming

QuantumClone (version 1.0.0.6)

parallelEM: Expectation Maximization algorithm

Description

Optimization of clone positions and proportion of mutations in each clone followed by filtering on most likely possibility for each mutation and a re-optimization. Then gives out the possibility with maximal likelihood Relies on foreach

Usage

parallelEM(Schrod, nclust, epsilon, contamination, prior_center = NULL,
  prior_weight = NULL, Initializations = 1, optim = "default",
  keep.all.models = FALSE)

Arguments

Schrod

A list of dataframes (one for each sample), generated by the Patient_schrodinger_cellularities() function.

nclust

Number of clones to look for (mandatory if prior_center or prior_weight are null)

epsilon

Stopping condition for the algorithm: what is the minimal tolerated difference of position or weighted between two steps

contamination

Numeric vector with the fraction of normal cells contaminating the sample

prior_center

Clone coordinates (from another analysis) to be used

prior_weight

Prior on the fraction of mutation in each clone

Initializations

Maximal number of independant initial condition tests to be tried

optim

use L-BFS-G optimization from R ("default"), or from optimx ("optimx"), or Differential Evolution ("DEoptim")

keep.all.models

Should the function output the best model (default; FALSE), or all models tested (if set to true)