MAMA (version 1.0.2)

compute.RQ: Function to compute R and Q statistics as defined in - see References

Description

Function computes R (average rank across studies) and Q (sum of the squared deviations of each study's rank for the gene from the mean of the ranks for that gene)

Usage

compute.RQ(RAN)

Arguments

RAN

matrix with rank of genes as produced by rank.genes, with rows coresponding to genes and columns coresponding to studies

Value

• matrix with first column of R statistic and second of Q statistic

References

Zintzaras, E., Ioannidis, J.P.A 2008 Meta-analysis for ranked discovery datasets: Theoretical framework and empirical demonstration for microarrays, Computational Biology and Chemistry 32, 39-47

See Also

rank.genes,MCtest

Examples

data(Singhdata)

#compute T-statistics and P-value
p1<-apply(Singhdata$esets[[1]],1,function(x) {t=t.test(x~Singhdata$classes[[1]], alternative="two.sided"); return(t$p.value) })
p2<-apply(Singhdata$esets[[2]],1,function(x) {t=t.test(x~Singhdata$classes[[2]], alternative="two.sided"); return(t$p.value) })
p3<-apply(Singhdata$esets[[3]],1,function(x) {t=t.test(x~Singhdata$classes[[3]], alternative="two.sided"); return(t$p.value) })
T1<-apply(Singhdata$esets[[1]],1,function(x) {t=t.test(x~Singhdata$classes[[1]], alternative="two.sided"); return(t$statistic) })
T2<-apply(Singhdata$esets[[2]],1,function(x) {t=t.test(x~Singhdata$classes[[2]], alternative="two.sided"); return(t$statistic) })
T3<-apply(Singhdata$esets[[3]],1,function(x) {t=t.test(x~Singhdata$classes[[3]], alternative="two.sided"); return(t$statistic) })

# Rank genes
rank1<-rank.genes(T1,p1)
rank2<-rank.genes(T2,p2)
rank3<-rank.genes(T3,p3)
# Compute R^star and Q^star 
RANK<-cbind(rank1,rank2, rank3)
RQ<-compute.RQ(RANK)
head(RQ)