ind.analysis: Identify differentially expressed genes in each individual...

In MetaDE: MetaDE: Microarray meta-analysis for differentially expressed gene detection

Description

ind.analysis is a function to perform individual analysis. The outputs are measures (p-values) for meta-analysis.

Usage

ind.analysis(x, ind.method = c("f", "regt", "modt", "pairedt",
                 "pearsonr", "spearmanr", "F", "logrank"), miss.tol =
                 0.3, nperm = NULL, tail, ...)

Arguments

x	a list of studies. Each study is a list with components: x: the gene expression matrix. y: the outcome variable. For a binary outcome, 0 refers to "normal" and 1 to "diseased". For a multiple class outcome, the first level being coded as 0, the second as 1, and so on. For survival data, it is the survial time of the paitents. censoring.status: 0 refers to individual who did not experimented the outcome while 1 is used for patients who develop the event of interest.
miss.tol	The maximum percent missing data allowed in any gene (default 30 percent).
nperm	The number of permutations. If nperm is NULL,the results will be based on asymptotic distribution.
ind.method	a character vector to specify the statistical test to test if there is association between the variables and the labels (i.e. genes are differentially expressed in each study). see "Details".
tail	a character string specifying the alternative hypothesis,must be one of "abs" (default), "low" or "high".
...	Additional arguments.

Details

The available statistical tests for argument, ind.method, are:

• "regt": Two-sample t-statistics (unequal variances).

• "modt": Two-sample t-statistics with the variance is modified by adding a fudging parameter. In our algorithm, we choose the penalized t-statistics used in Efron et al.(2001) and Tusher et al. (2001). The fudge parameter s0 is chosen to be the median variability estimator in the genome.

• "pairedt": Paired t-statistics for the design of paired samples.

• "pearsonr": Pearson's correlation. It is usually chosen for quantitative outcome.

• "spearmanr": Spearman's correlation. It is usually chosen for quantitative outcome.

• "F": the test is based on F-statistics. It is usually chosen where there are 2 or more classes.

For the argument, miss.tol, the default is 30 percent. For those genes with less than miss.tol *100 percent missing are imputed using KNN method in package,impute; for those genes with more than or equal miss.tol*100 percent missing are igmored for the further analysis.

Value

a list with components:

stat	the value of test statistic for each gene
p	the p-value for the test for each gene
bp	the p-value from nperm permutations for each gene. It will be used for the meta analysis. It can be NULL if you chose asymptotic results.

References

Jia Li and George C. Tseng. (2011) An adaptively weighted statistic for detecting differential gene expression when combining multiple transcriptomic studies. Annals of Applied Statistics. 5:994-1019.

Shuya Lu, Jia Li, Chi Song, Kui Shen and George C Tseng. (2010) Biomarker Detection in the Integration of Multiple Multi-class Genomic Studies. Bioinformatics. 26:333-340. (PMID: 19965884; PMCID: PMC2815659)

See Also

MetaDE.Read,MetaDE.match,MetaDE.merge,MetaDE.filter,MetaDE.pvalue and MetaDE.rawdata

Examples

#--generate two pseudo datasets----#
label1<-rep(0:1,each=5)
label2<-rep(0:1,each=5)
exp1<-cbind(matrix(rnorm(5*20),20,5),matrix(rnorm(5*20,2),20,5))
exp2<-cbind(matrix(rnorm(5*20),20,5),matrix(rnorm(5*20,1.5),20,5))

#the input has to be arranged in lists
x<-list(list(x=exp1,y=label1),list(x=exp2,y=label2))

# start individual analysis for each study: 
#find genes whose expession is higher in class 2 vs class 1 using moderated t test for both studies
test1<-ind.analysis(x,ind.method=c("modt","modt"),tail="high",nperm=100) 
#here I want to use two-sample t test for study 1 and moderated t test for study 2.
test2<-ind.analysis(x,ind.method=c("regt","modt"),tail="abs",nperm=100)

#--------time to event---------#
#--generate three pseudo datasets----#
exp1<-matrix(rnorm(20*10),20,10)
time1=c(4,3,1,1,2,2,3,10,5,4)
event1=c(1,1,1,0,1,1,0,0,0,1)
#study 2
exp2<-matrix(rnorm(20*10,1.5),20,10)
time2=c(4,30,1,10,2,12,3,10,50,2)
event2=c(0,1,1,0,0,1,0,1,0,1)
#study 3
exp3<-matrix(rnorm(20*15),20,15)
time3=c(1,27,40,10,2,6,1,10,50,100,20,5,6,8,50)
event3=c(0,1,1,0,0,1,0,1,0,1,1,1,1,0,1)

#the input has to be arranged in lists
test3<-list(list(x=exp1,y=time1,censoring.status=event1),list(x=exp2,y=time2,censoring.status=event2),
list(x=exp3,y=time3,censoring.status=event3))

# start individual analysis for each study: i use log rank test for all studies
test3.res<-ind.analysis(test3,ind.method=rep("logrank",3),nperm=100,tail='abs')
 

MetaDE documentation built on May 29, 2017, 9 a.m.