Datasets For Multiple Hypothesis Testing

Question

[Kami]

Hello,

I would like to ask you if you know where I can find a large dataset describing the following experiment: I have two treatments each given to two groups of people and I want to get observations for different therapeutical properties for the two groups. That is I am looking for datasets for multiple endpoints problem in order to apply the theory about multiple hypothesis testing.

Thank you, Kami

[User 59]

What kind of dataset? Arrays? Proteomics? Metabolomics? RNA-Seq?

[Kami]

0

I want to use R in order to analyse the dataset; I think that arrays would be better.

Answer 1

[Chris Evelo]

I guess then it would be best to search the large microarray data repositories [?]Arrayexpress[?] and [?]GEO[?].

Answer 2

[David W]

Kami,

You talked about R packages inlcuding multcomp. These packages often come with example datasets that you can use to get to grips with how the methods work. The dataset mtept from multcomp sounds like the sort of thing you're after (2 treatments, 4 endpoints)

library(multcomp)
data(mtept)
str(mtept)

#'data.frame':   111 obs. of  5 variables:
# $ treatment: Factor w/ 2 levels "Drug","Placebo": 2 2 2 2 2 2 2 2 2 2 ...
# $ E1       : int  4 5 1 4 3 4 2 2 3 2 ...
# $ E2       : int  3 0 0 0 0 1 0 2 0 0 ...
# $ E3       : int  3 1 1 3 2 2 4 5 1 1 ...
# $ E4       : int  5 7 9 5 9 6 6 5 7 9 ...

It doesn't have the patients split into two groups, but you can make that up:

grp <- as.factor(c(rep(c('one', 'two'), 50), 'two'))
head(grp)
# [1] one two one two one two
# Levels: one two

[Kami]

Do you also have in mind any other datasets like this, but with more endpoints? That would be great. Thank you.

Answer 3

[Kami]

I want to use R in order to analyse the dataset; I think that arrays would be better.

Answer 4

[Kami]

Actually, anything may help, I have to try it in R. I am interested to give an example of how multiple hypothesis testing works so I don't mind what kind of dataset would be.

Also do you know where I can look for some examples of how we analyse datasets in R (regarding multiple hypothesis testing)? I have just find different packages like multtest, multcomp that apply in my case but still i need data to work on them.

[David W]

Hi Kami, with "question and answer" type sites like the stackexchange ones you should try use the the "your answer" section for answers to the original question. Clarifications like this should be added as edits to the original questions or comments underneath it.