ISOpureS1.model_optimize.opt_mm: Optimize the reference cancer profile, m, in step 1

In ISOpureR: Deconvolution of Tumour Profiles

Description

The goal of this function is to optimize the reference cancer profile mm. Because mm is constrained (must be parameters of multinomial/discrete distribution), we don't directly optimize the likelihood function w.r.t. mm, but we perform change of variables to do unconstrained optimization. We therefore store these unconstrained variables in the field "mm_weights", and update these variables.

Usage

ISOpureS1.model_optimize.opt_mm(
	tumordata, model, 
	NUM_ITERATIONS_RMINIMIZE, 
	iter, 
	NUM_GRID_SEARCH_ITERATIONS
	) 

Arguments

tumordata	a GxD matrix representing gene expression profiles of tumour samples
model	list containing all the parameters to be optimized
NUM_ITERATIONS_RMINIMIZE	minimum number of iteration that the minimization algorithm runs
iter	the iteration number
NUM_GRID_SEARCH_ITERATIONS	number of times to try restarting with different initial values

Value

The model with mm_weights updated (and log_all_rates)

Author(s)

Gerald Quon, Catalina Anghel, Francis Nguyen

ISOpureR documentation built on May 11, 2019, 1:02 a.m.