MEET (version 5.1.1)

kfold.PCA: PCA

Description

For a vector of principal components the Q-residuals of a DNA symbolical sequence are calculated using a PCA model of the numerical DNA motif. First sequences are converted to numerical DNA sequences and then the model is applied. A matrix is returned for each number of PC with the value of the Q-residuals in each position and a label indicating if the sequence belong to a binding site or not

Usage

kfold.PCA(iicc, TF)

Arguments

iicc

Options described in the MEET function

TF

DNA sequences used to construct the motif model

Value

As a list, for each nPCs

matriuROC[[nPCs]]

matrix with two colums, in the first one the Q-residuals for each studied sequence, and the second one indicates if the sequence belong to a TFBS

Details

Alignment method has to be installed in your computer.

References

Jolliffe I.T. Principal Component Analysis, Series: Springer Series in Statistics, 2nd ed., Springer, NY, 2002, XXIX, 487 p. 28 illus. ISBN 978-0-387-95442-4 Stacklies, Wolfram, Redestig, Henning, Scholz, Matthias, Walther, Dirk, and Selbig, Joachim: pcaMethods a bioconductor package providing PCA methods for incomplete data, Bioinformatics 23(9), volume 23, 1164-1167, 2007

See Also

MEET, kfold.Entropy, kfold.MATCH, kfold.Divergence, PCanalysis, llegir_DNA, convertDNA, numericalDNA.

Examples

require("MEET")
require("seqinr")
pathMEET <- system.file("sequences", package = "MEET")
data(iicc)
data(TranscriptionFactor)
iicc$method<-"PCA"
#Define the numner of principal components
iicc$vector<-c(1,3,5)
kfold.PCA(iicc, TF = paste(pathMEET, "AP1.fa", sep = "/"))