| topGene {RankProd} | R Documentation |
Identify differentially expressed genes using rank product method
topGene(x,cutoff=NULL,method="pfp",num.gene=NULL,logged=TRUE,logbase=2,gene.names=NULL)
x |
the value returned by the function RP, RP.advance or Rsum.advance |
cutoff |
threshold in pfp used to select genes |
method |
If cutoff is provided, the method needs to be selected to identify genes."pfp" uses percentage of false prediction, which is the default setting. "pval" used p-value which is less stringent than pfp |
num.gene |
number of candidate genes of interests, if cutoff is provided, this will be ignored |
logged |
if "TRUE", data has bee logged, otherwise set it to "FALSE" |
logbase |
base used when taking log, used to restore the fold change.The default value is 2, this will be ignored if logged=FALSE |
gene.names |
if "NULL", no gene name will be attached to the output table |
Two tables of identified genes with
gene.index: index of gene in the original data set
RP/Rsum: Computed rank product/sum for each gene
FC:(class1/class2): Expression Fold change of class 1/ class 2.
pfp: estimated pfp for each gene if the gene is used as cutoff point
P.value: estimated p-value for each gene
Table 1 list genes that are up-regulated under class 2, Table 1 ist
genes that are down-regulated under class 2,
Fangxin Hong fhong@salk.edu
Breitling, R., Armengaud, P., Amtmann, A., and Herzyk, P.(2004) Rank Products: A simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments, FEBS Letter, 57383-92
# Load the data of Golub et al. (1999). data(golub)
# contains a 3051x38 gene expression
# matrix called golub, a vector of length called golub.cl
# that consists of the 38 class labels,
# and a matrix called golub.gnames whose third column
# contains the gene names.
data(golub)
#use a subset of data as example, apply the rank
#product method
subset <- c(1:4,28:30)
#Setting rand=123, to make the results reproducible,
#identify genes
RP.out <- RP(golub[,subset],golub.cl[subset],rand=123)
#get two lists of differentially expressed genes
#by setting FDR (false discivery rate) =0.05
table=topGene(RP.out,cutoff=0.05,method="pfp",logged=TRUE,logbase=2,
gene.names=golub.gnames[,3])
table$Table1
table$Table2
#using pvalue<0.05
topGene(RP.out,cutoff=0.05,method="pval",logged=TRUE,logbase=2,
gene.names=golub.gnames[,3])
#by selecting top 10 genes
topGene(RP.out,num.gene=10,gene.names=golub.gnames[,3])