mogsa (version 1.6.4)

mbpca: Extension of PCA to analyze multiple data sets

Description

Three approaches are supplied in this function, consensus PCA (CPCA), generalized CCA (GCCA) and multiple co-inertia analsyis (MCIA).

Usage

mbpca(x, ncomp, method, k = "all", center = TRUE, scale = FALSE,  option = "uniform", maxiter = 1000, moa = TRUE, verbose = TRUE,  svd.solver = c("svd", "fast.svd", "propack"))

Arguments

x

A list of matrix or data.frame, where rows are variables and columns are samples. The columns among the matrices need to be match but the variables do not need to be.

ncomp

An integer; the number of components to calculate. To calculate more components requires longer computational time.

method

A character string could be one of c("globalScore", "blockScore", "blockLoading"). The "globalScore" approach equals consensus PCA; The "blockScore" approach equals generalized canonical correlation analysis (GCCA); The "blockLoading" approach equals multiple co-inertia anaysis (MCIA);

k

The absolute number (if k >= 1) or the proportion (if 0

center

Logical; if the variables should be centered

scale

Logical; if the variables should be scaled

option

A charater string could be one of c("lambda1", "inertia", "uniform") to indicate how the different matrices should be normalized. If "lambda1", the matrix is divided by its the first singular value, if "inertia", the matrix is divided by its total inertia (sum of square), if "uniform", none of them would be done.

maxiter

Integer; Maximum number of iterations in the algorithm

moa

Logical; whether the output should be converted to an object of class moa-class

verbose

Logical; whether the process (# of PC) should be printed

svd.solver

A charater string could be one of c("svd", "fast.svd", "propack"). The default "fast.svd " has a good compromise between the robustness and speed. "propack" is the fastest but may failed to converge in practice.

Value

An object of class moa-class (if moa=TRUE) or an list object contains the following elements:tb - the block scorespb - the block loadingst - the global scoresw - the wegihts of block scores to construct the global scor

Details

details need to update

References

reference need to be updated

See Also

see moa for non-iterative algorithms for multi-block PCA.

Examples

data("NCI60_4arrays")
tumorType <- sapply(strsplit(colnames(NCI60_4arrays$agilent), split="\\."), "[", 1)
colcode <- as.factor(tumorType)
levels(colcode) <- c("red", "green", "blue", "cyan", "orange", 
                     "gray25", "brown", "gray75", "pink")
colcode <- as.character(colcode)



moa <- mbpca(NCI60_4arrays, ncomp = 10, k = "all", method = "globalScore", option = "lambda1", 
             center=TRUE, scale=FALSE)
plot(moa, value="eig", type=2)
r <- bootMbpca(moa, mc.cores = 1, B=6, replace = FALSE, resample = "sample")

moas <- mbpca(NCI60_4arrays, ncomp = 3, k = 0.1, method = "globalScore", option = "lambda1", 
              center=TRUE, scale=FALSE)


scr <- moaScore(moa)
scrs <- moaScore(moas)
diag(cor(scr[, 1:3], scrs))

layout(matrix(1:2, 1, 2))
plot(scrs[, 1:2], col=colcode, pch=20)
legend("topright", legend = unique(tumorType), col=unique(colcode), pch=20)
plot(scrs[, 2:3], col=colcode, pch=20)

gap <- moGap(moas, K.max = 12, cluster = "hcl")
gap$nClust


hcl <- hclust(dist(scrs))
cls <- cutree(hcl, k=4)
clsColor <- as.factor(cls)
levels(clsColor) <- c("red", "blue", "orange", "pink")
clsColor <- as.character((clsColor))

heatmap(t(scrs[hcl$order, ]), ColSideColors = colcode[hcl$order], Rowv = NA, Colv=NA)
heatmap(t(scrs[hcl$order, ]), ColSideColors = clsColor[hcl$order], Rowv = NA, Colv=NA)

genes <- moaCoef(moas)
genes$nonZeroCoef$agilent.V1.neg