runCBN: Runs H-CBN

In cbg-ethz/pathTiMEx: Joint Inference of Mutually Exclusive Driver Pathways and their Progression Dynamics

Description

runCBN runs the H-CBN program from the the command line, for already existing .pat (binary alteration matrix) and .poset (initial solution for structure) files.

Usage

runCBN(path, name, optionsSA, noThreads)

Arguments

path	path where the folder containing the initial results is created, as well as where the final results are written. In the H-CBN help file, this is the path where the filestem directory is created.
name	name of the directory where results are initially written.
optionsSA	string representing the command-line options of the simmulated annealing optimization routine.
noThreads	number of threads on which to run H-CBN.

Details

The folder were results of the H-CBN run are initially saved is completeFilestem=path/name (which, if not existent, is created). After running this function, five files are written: the optimized structure (poset), written to path/name/00000.poset and also to path/name.poset, replacing the initial poset and becoming the initial solution for the next iteration of the simmulated annealing optimization; the optimized lambda parameters, written to path/name.lambda and also to path/name/paramsEps.lambda, where it also includes the estimate of the error-rate epsilon; and the log of the simmulated annealing repetitions, which are written to path/name.log. The returned optimal structure contains an additional dummy row and column identically set to 0, in preparation for the format required by the local group optimization.

Value

list consisting of

• optStruct a square matrix of dimension input number of pathways plus 1, representing the inferred optimal structure. The last row and column correspond to a dummy pathway and are identically zero.

• eps H-CBN estimate for the gene-wise error rate (positive real number)

• alpha H-CBN estimate for alpha (positive real number).

• loglik H-CBN estimate for the log likelihood.

• lamobs H-CBN estimate for the waiting time rate of the observation time (fixed to 1).

• lams vector, H-CBN estimate for the waiting time rates of the events (pathways).

Author(s)

Simona Constantinescu, [e-mail]

cbg-ethz/pathTiMEx documentation built on May 13, 2019, 2:03 p.m.