similar to: FDR for hypergeometric tests

I am posting this to both R and BioC communities because I believe there is a lot of confusion on this topic in both communities (having searched the mail archives of both) and I am hoping that someone will have information that can be shared with both communities. I have seen countless questions on the BioC list regarding limma (Bioconductor) and its calculation of FDR. Some of them involved

I am posting this to both R and BioC communities because I believe there is a lot of confusion on this topic in both communities (having searched the mail archives of both) and I am hoping that someone will have information that can be shared with both communities. I have seen countless questions on the BioC list regarding limma (Bioconductor) and its calculation of FDR. Some of them involved

Dear All I have a question about the hypergeomteric distribution. Example 1: I have a universe of 6187 objects, and 164 have a particular attribute, therefore 6187-164 do not have that attribute. I sample 249 of those objects, and find that 19 have that attribute. I get a p-value here (looking at just over-representation): phyper(19, 164, 6187-164, 249, lower.tail=FALSE) [1] 7.816235e-06

Mark, there is a fdr website link via Yoav Benjamini's homepage which is: http://www.math.tau.ac.il/%7Eroee/index.htm On it you can download an S-Plus function (under the downloads link) which calculates the false discovery rate threshold alpha level using stepup, stepdown, dependence methods etc. Some changes are required to the plotting code when porting it to R. I removed the

In both versions of R to which I currently have access (R-0.16.1 and R-0.61.1), "phyper" stops returning correct cumulative probabilities as the parameters of the hypergeometric distribution get large. For example, when N1=1345, N2=1055, and n=1330, phyper returns either 0 or 1, and nothing in between. Looking at phyper.c, it's clear what's happening. First a term (called

Dear R Users, I employed the phyper() function to estimate the likelihood that the number of genes overlapping between 2 different lists of genes is due to chance. This appears to work appropriately. Now i want to try this with 3 lists of genes which phyper() does not appear to support. Some googling suggests i can utilize the Multivariate hypergeometric distribution to achieve this. eg.:

You asked the same question on the Bioconductor mailing list back in August. At that time, you suggested yourself a solution for how the adjusted p-values should be interpreted. I answered your query and told you that your interpretation was correct. So I'm not sure what more can be said, except that you should read the article Wright (1992), which is cited in the help entry for p.adjust(),

Dear R team, I have a simple question. I tried this command: phyper(17,449,19551,181, FALSE) [1] 1.47295e-07 and then I tried this command: (fisher.test(matrix(c(17,449,181,19551),2,2), alternative='greater'))$p.value [1] 3.693347e-06 Shouldn't be identical the results of the two commands ? What is the difference ? Thx a lot -- View this message in context:

Hi, I hope somebody can help me on how to use the hypergeometric function. I did read through the R documentation on hypergeometric but not really sure what it means. I would like to evaluate the hypergeometric function as follows: F((2*alpha+1)/2, (2*alpha+2)/2 , alpha+1/2, betasq/etasq). I'm not sure which function should be used- either phyper or qhyper or dhyper Where

Hello! I use R-version 1.0.0 To get the 0.95 quantile of the hypergeometric distribution with the parameters m=45000,n=5000 and k=600 I use the R-command > qhyper(0.95,45000,5000,600). The value obtained is 600. However, the true value is 552. The latter can be obtained for example by calling the corresponding distribution function with the R commands > x<-540:580 >

Hola a todos, Tengo un pequeño problemilla... Tengo unas 9000 variables que he contrastado con 1 en concreto con el test de wilcoxon. He calculado el p-valor, y queria corregirlo con el permutation-based FDR. He encontrado una funcion con R comp.fdr()que hace esta corrección, pero te pide que le pongas las variables con las observaciones y te hace el test (según he entendido). Yo solo quiero

To All How does one get a list of functions from a loaded package so that one can then get the appropriate help for each of the functions. Currently my method is based on a lot of trial-and-error. Here's an example of what I mean... >From this forum I learn that an interesting package called "multtest" exists on Bioconductor. I then use R Console's "Packages" --

You can get help on the whole package by > help(package="multtest") which is likely pretty close to what you want. There's the index etc for the package on the web as well. You can also just look in the package's installation directory. If it's loaded you can do an ls(2) say if it loaded in position 2, to get all objects in the package. Reid -----Original Message-----

Hello, I am not sure about the p.adjust( , fdr). How do these adjusted p-values get? I have read papers of BH method. For independent case, we compare the ordered p-values with the alfa*i/m, where m is the number of tests. But I have checked that result based on the adjusted p-values is different with that by using the independent case method. Then how do the result of p.adjust( , fdr) come? And

Aloha all, I know too little about what I'm about to write and hope I'm not wasting your time. For a class I'm teaching in archaeological data analysis, I'm trying to put together a routine that calculates the so-called Petersen index and, especially, confidence intervals for the index. This was introduced to archaeologists by N.R.J. Fieller and A. Turner in an article

The Partek package (www.partek.com) allows only two selections for Multiple Test Correction: Bonferroni and Dunn-Sidak. Can anyone suggest why Partek implemented Dunn-Sidak and not the other methods that R has? Is there any particular advantage to the Dunn-Sidak method? R knows about these methods (in R 2.1.1): > p.adjust.methods [1] "holm" "hochberg" "hommel"

Hello R experts, I am trying to analyze this dataset and am stuck on this problem for quite some time now. I am using mt.rawp2adjp. the output that came out was a matrix with two colums since I had asked it to calculate the adjusted p values using one method. so it has the two columns as: rawp BH I combined these using cbind with my actual dataframe. checked using head all was fine. thereafter I

> dim(data.sub) [1] 10000 140 #####extracting all differentially express genes########## library(multtest) two_side<- (1-pt(abs(data.sub),50))*2 diff<- mt.rawp2adjp(two_side) all_differ<-diff[[1]][37211:10000,] all_differ #####list of differentially expressed genes########## > probe.names<- + all_differ[[2]][all_differ[[1]][,"BY"]<=0.01] Error in

Hi all, I have tried to compute a p-value for a hypergeometric distribution as: dhyper(x,k,l,n) + phyper(x,k,l,n,lower.tail=FALSE) and sometimes obtained negative values. Do you know if it is because a round error or am I doing something wrong? Thanks in advance, Aurora

dear all, I recently came across the following issue and I was not sure whether it is intentionally or not: using p.adjust to adjust p-values for multiple hypothesis testing using the method from Benjamini and Hochberg removes all NA values from the input vector and does not account for them in the adjustment, i.e. in a vector of 23 p-values with 20 of them being NA it adjusts the 3 non-NA