[BioC] more multtest problems

[michael watson (IAH-C)]

Hi

Even after some very good help, I am still confused by multtest.

Below is the output from my analysis.  What I basically do is read in some data, use t.test() to perform a two-tailed, unequal variance t-test on the first row, then use mt.maxT() to perform a t-test on all the rows and then compare the results.

What I have found is that although t.test() and mt.maxT() produce the same t value (0.9796), the raw p-values differ (0.3954 vs 0.54285714).

Why is rawp from mt.maxT different to the p-value from t.test()

Thanks!

Mick


> library(multtest)
> data <- read.table("average.txt", header=TRUE, sep="\t", quote="\"")
> cl = c(1,1,1,1,0,0,0)
> # do two-tailed t-test, unequal variance
> t <- t.test(data[1,1:4],data[1,5:7], alternative="two.sided", var.equal=FALSE)
> t

        Welch Two Sample t-test

data:  data[1, 1:4] and data[1, 5:7]
t = 0.9796, df = 3.199, p-value = 0.3954
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -2.407296  4.659662
sample estimates:
  mean of x   mean of y
1.130890500 0.004707667

> multt <- mt.maxT(data,cl,test="t",side="abs")
We'll do complete enumerations

We're doing 35 complete permutations
b=1     b=2     b=3     b=4     b=5     b=6     b=7     b=8     b=9     b=10
b=11    b=12    b=13    b=14    b=15    b=16    b=17    b=18    b=19    b=20
b=21    b=22    b=23    b=24    b=25    b=26    b=27    b=28    b=29    b=30
b=31    b=32    b=33    b=34    b=35    >
> multt
  index  teststat       rawp       adjp
1     4 2.4344730 0.02857143 0.02857143
2     3 1.7019798 0.25714286 0.25714286
3     5 1.5719193 0.08571429 0.25714286
4     6 1.3674557 0.02857143 0.25714286
5     2 1.1519341 0.20000000 0.28571429
6     1 0.9796023 0.54285714 0.54285714