R/p.vector.R

In maSigPro: Significant Gene Expression Profile Differences in Time Course Gene Expression Data

p.vector <- function (data, design, Q = 0.05, MT.adjust = "BH", min.obs = 6, counts=FALSE, family=NULL, theta=10, epsilon=0.00001, item="gene") 
{
    if (is.data.frame(design) || is.matrix(design)) {
        dis <- design
        groups.vector = NULL
        edesign = NULL
    }
    else if (is.list(design)) {
        dis <- as.data.frame(design$dis)
        groups.vector <- design$groups.vector
        edesign <- design$edesign
    }

    if (is.null(family)) {
	if(!counts) { family=gaussian() }
	if(counts)  { family=negative.binomial(theta) }
    }

    dat <- as.matrix(data)
    dat <- dat[, as.character(rownames(dis))]
    G <- nrow(dat)

# Removing rows with many missings:

    count.na <- function(x) (length(x) - length(x[is.na(x)]))
    dat <- dat[apply(dat, 1, count.na) >= min.obs, ]

# Removing rows with all ceros:

    sumatot <- apply(dat, 1, sum)
    counts0 <- which(sumatot == 0)
    if (length(counts0) > 0) { dat <- dat[-counts0,] } 

    g <- dim(dat)[1]
    n <- dim(dat)[2]
    p <- dim(dis)[2]
    p.vector <- vector(mode = "numeric", length = g)

#----------------------------------------------------------------------
    for (i in 1:g) {

        y <- as.numeric(dat[i, ])

        div <- c(1:round(g/100)) * 100
        if (is.element(i, div)) 
            print(paste(c("fitting ",item, i, "out of", g), collapse = " "))

        model.glm<- glm(y~.,data=dis , family=family, epsilon=epsilon)
	  if(model.glm$null.deviance==0) { p.vector[i]=1 } else{

        model.glm.0<-glm(y~1, family=family, epsilon=epsilon)

        if(family$family=="gaussian")
        {
                test<-anova(model.glm.0,model.glm,test="F")
                if( is.na(test[6][2,1]) ) {p.vector[i]=1}
                else{ p.vector[i]=test[6][2,1]}
        }
        else
        {
                test<-anova(model.glm.0,model.glm,test="Chisq")
                if( is.na(test[5][2,1]) ) {p.vector[i]=1}
                else{ p.vector[i]=test[5][2,1]}
        }
 }
}

#----------------------------------------------------------------------
    p.adjusted <- p.adjust(p.vector, method = MT.adjust, n = length(p.vector))
    genes.selected <- rownames(dat)[which(p.adjusted <= Q)]

    FDR <- sort(p.vector)[length(genes.selected)]
   
    SELEC <- as.matrix(as.data.frame(dat)[genes.selected, ])
      if (nrow(SELEC) == 0) 
        print("no significant genes")

    p.vector <- as.matrix(p.vector)
    rownames(p.vector) <- rownames(dat)
    colnames(p.vector) <- c("p.value")

#-------------------------------------------------------------------------

    output <- list(SELEC, p.vector, p.adjusted, G, g, FDR, 
        nrow(SELEC), dis, dat, min.obs, Q, groups.vector, edesign, family)
    names(output) <- c("SELEC", "p.vector", "p.adjusted", "G", 
        "g", "FDR", "i", "dis", "dat", "min.obs", "Q", "groups.vector", 
        "edesign", "family")
    output
}