[BioC] goseq: biased distribution of p-values

[Joachim Jacob]

Hi all,


I have run goseq on RNA-seq data differential expression results, but 
got a strange p-value distribution: from zero to almost one it is 
uniformly distributed, but there is a huge enrichment to the value one. 
See screenshot: https://dl.dropbox.com/u/18352887/Selection_042.png

This seems that the distribution underlying the null hypothesis is 
incorrect? As a consequence, nearly all conditions I test do not have 
overrepresented GO categories. I do not trust the outcome of this.

I've done goseq analysis with custom GO-mappings (from Biomart) and 
lengths file (from GenomicFeatures package) for the genome sacCer3 (gene 
database sgdGene). (when I tried the built-in genome, goseq failed to 
find the data automatically).

Any help on how to interpret or to correct this?


Thanks,
Joachim

Joachim Jacob, PhD

Rijvisschestraat 120, 9052 Zwijnaarde
Tel: +32 9 244.66.34
Bioinformatics Training and Services (BITS)
http://www.bits.vib.be
@bitsatvib

On 10/30/2012 08:53 AM, Dave Tang wrote:
> Hello,
>
> I've been trying to install goseq on R-2.15.2 but one of the 
> dependencies, RCurl is not installing. I've compiled curl and 
> curl-config (7.28.0); there is also a system wide installation of curl 
> (7.15.5) in /usr/bin/. I tried installing just RCurl but to no avail. 
> I've looked online for help for hours but it has been hopeless. I 
> would be extremely grateful if someone could let me know what is wrong.
>
> source("http://bioconductor.org/biocLite.R")
> biocLite("RCurl")
> checking for curl-config... /home/davetang/bin/bin/curl-config
>
> [snipped]
>
> gcc -std=gnu99 -shared -L/usr/local/lib64 -o RCurl.so base64.o curl.o 
> curlInit.o curl_base64.o enums.o json.o memoryManagement.o myUTF8.o 
> -L/home/davetang/bin/curl/lib -lcurl -lxml2 -lz -lm
> installing to /home/davetang/src/R-2.15.2/library/RCurl/libs
> ** R
> ** data
> ** inst
> ** preparing package for lazy loading
> Creating a generic function for 'close' from package 'base' in package 
> 'RCurl'
> ** help
> *** installing help indices
> ** building package indices
> ** installing vignettes
> ** testing if installed package can be loaded
> Error in dyn.load(file, DLLpath = DLLpath, ...) :
>   unable to load shared object 
> '/home/davetang/src/R-2.15.2/library/RCurl/libs/RCurl.so':
>   libcurl.so.4: cannot open shared object file: No such file or directory
> Error: loading failed
> Execution halted
> ERROR: loading failed
> * removing '/home/davetang/src/R-2.15.2/library/RCurl'
>
> The downloaded source packages are in
>         '/tmp/Rtmp9heDvL/downloaded_packages'
> Updating HTML index of packages in '.Library'
> Making packages.html  ... done
> Warning message:
> In install.packages(pkgs = pkgs, lib = lib, repos = repos, ...) :
>   installation of package 'RCurl' had non-zero exit status
>
> sessionInfo()
> R version 2.15.2 (2012-10-26)
> Platform: x86_64-unknown-linux-gnu (64-bit)
>
> locale:
> [1] C
>
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods base
>
> other attached packages:
> [1] BiocInstaller_1.8.3
>
> loaded via a namespace (and not attached):
> [1] tools_2.15.2
>
> Many thanks in adavance,
>