cnmtf: cNMTF algorithm

In lgl15/cnmtf: Prioritisation of single nucleotide variants using Corrected non-negative matrix tri-factorisation

Description

Function to perform one repetition of cNMTF

Usage

cnmtf(R, HLH, Vo, Wu, k, iters = 1, calcObj, calcObj2, displ = TRUE,
  tof = 1e-05, lparameters, init = 1, V.init = NULL, U.init = NULL,
  init.S = FALSE, DWD = NULL)

Arguments

R	Data matrix, n x m
HLH	Term in the correction for population structures. Dot product: H . L . H where: H is the Centering matrix (m x m) and L is Ancestry matrix (i.e. Kernel on the confounder random variable) (m x m)
Vo	Phenotype/Outcome label matrix, m x k2
Wu	Adjacency matrix of the SNV-SNV network. n x n
k	Vector of rank parameters, k1 x k2
iters	Default number of itersatiors
calcObj	Check convergency each X number of itersations
displ	Logical. Print number of iterations
tof	RelativeError
lparameters	Vector of regularization parameters: γ_{1}, γ_{2}, γ_{3} for the SNV-SNV network, the Phenotype matrix and the Ancestry Kernel respectively.
init	Initialize the matrices randomly (0) or by using PSVD (1)
DWD	Term in the normalised graph laplacian. Dot product: 1/sqrt(Du) . Wu . 1/sqrt(Du), where: Du is the Degree matrix and Wu is the adjacency matrix of the SNV-SNV network.

Value

• U low-dimensional cluster indicator matrix for features observed data, n x k1

• V low-dimensional cluster indicator matrix for observed data, m x k2

• S cluster mapping matrix, k1 x k2

• num.iters number of itersation till convergence

• final.J Final objective function value

• objF.vector Objective function values across iterations

Author(s)

Luis G. Leal, [e-mail]

References

• Shang, Fanha, et al. "Graph dual regularization non-negative matrix factorization for co-clustering" Pattern Recognition 45 (2012)

• Li, Rakitisch, et al. "ccSVM: correcting SVMs for confounding factors in biological data classification" ISMB 2011

See Also

Other Factorisation functions: clus.membership, consensus.clust, hierarchical.clust, initialise.UV, neg.constrain, parameters.cnmtf, plot.parameter, pos.constrain, psvd.init, regression.snps, score.cnmtf, synthetic.gwas

lgl15/cnmtf documentation built on May 28, 2019, 6:33 p.m.