BioinformaticsFMRP / TCGAbiolinksGUI

TCGAbiolinksGUI code development for R/Bioconductor submission http://tcgabiolinks.fmrp.usp.br:3838/
http://bioinformaticsfmrp.github.io/TCGAbiolinksGUI
35 stars 14 forks source link

TCGAbiolinksGUI ELMER visualization fails: ERROR: [on_request_read] connection reset by peer > #40

Open yinlisssss opened 6 years ago

yinlisssss commented 6 years ago

Hi, I really love this package, which helps me a lot.

recently I have been trying to learn how to perfome the ELMER analysis, the first 2 steps(create input data and analysis) worked very well, and I ended up having this file called permu.rda, then I tried to do the third step(visualize results), but as long as I input permu.rda by SELECT results, the R said ERROR: [on_request_read] connection reset by peer, then the website disappered.

thanks a lot if you could help me with this..

tiagochst commented 6 years ago

Hello, Could you send me the MAE object and the parameters you used? So I can replicate the case.

The permu.rda is not the final result expected that's why it is closing. I'll have to add a tryCatch. But there should be another .rda with all the results. Unless there no master TF regulator found.

yinlisssss commented 6 years ago

1 2 3 4

yinlisssss commented 6 years ago

the mae object is bigger than 10mb, how can i upload? thanks

tiagochst commented 6 years ago

Dropbox or google drive would work.

yinlisssss commented 6 years ago

OK, uploading to dropbox... the internet sucks, sorry for your time

yinlisssss commented 6 years ago

https://www.dropbox.com/s/zy1bk5r031zhvkp/ELMER.zip?dl=0

tiagochst commented 6 years ago

I was able to get results with the same parameters.

screen shot 2018-04-23 at 12 38 20 am

It seems the version you have is not the last version. Could you please update the package?

devtools::install_github("tiagochst/ELMER")
devtools::install_github("BioinformaticsFMRP/TCGAbiolinksGUI")

My only concern is the Empirical P value cut-off. We normally set Number of permutations to 10K to get a reliable correction. However this would take several hours to run. If you are doing only 100 permutations please reduce the Empirical P value cut-off to 0.05.

Also. why do you have that few number of probes?

tiagochst commented 6 years ago

Results file: https://drive.google.com/open?id=1rmBrMmykC7Fvoa6BNy7CiJ6DceVft1pZ

yinlisssss commented 6 years ago

I tried to install the latest version, however

ERROR: dependencies 'TCGAbiolinksGUI.data', 'IlluminaHumanMethylation450kanno.ilmn12.hg19', 'IlluminaHumanMethylation450kmanifest', 'IlluminaHumanMethylation27kmanifest', 'IlluminaHumanMethylation27kanno.ilmn12.hg19', 'IlluminaHumanMethylationEPICanno.ilm10b2.hg19', 'IlluminaHumanMethylationEPICmanifest' are not available for package 'TCGAbiolinksGUI'

yinlisssss commented 6 years ago

1 2 Then i manually downloaded last version ZIP from bioconductor . now it works.

but when it comes to ELMER analysis, it's like this..

tiagochst commented 6 years ago

Can you try running:

source("https://bioconductor.org/biocLite.R")
deps <- c("devtools",
                  "IlluminaHumanMethylation450kanno.ilmn12.hg19",
                  "IlluminaHumanMethylation450kmanifest", 
                  "IlluminaHumanMethylation27kmanifest", 
                  "IlluminaHumanMethylation27kanno.ilmn12.hg19", 
                  "IlluminaHumanMethylationEPICanno.ilm10b2.hg19",
                  "IlluminaHumanMethylationEPICmanifest")
for(pkg in deps)  if (!pkg %in% installed.packages()) biocLite(pkg, dependencies = TRUE)
devtools::install_github("tiagochst/ELMER.data")
devtools::install_github("BioinformaticsFMRP/TCGAbiolinksGUI.data",ref = "R_3.4")
devtools::install_github("BioinformaticsFMRP/TCGAbiolinksGUI")
yinlisssss commented 6 years ago

Downloading GitHub repo BioinformaticsFMRP/TCGAbiolinksGUI@master from URL https://api.github.com/repos/BioinformaticsFMRP/TCGAbiolinksGUI/zipball/master Installing TCGAbiolinksGUI "C:/PROGRA~1/R/R-34~1.4/bin/x64/R" --no-site-file --no-environ --no-save \ --no-restore --quiet CMD INSTALL \ "C:/Users/asus/AppData/Local/Temp/Rtmpa2S620/devtools29d4363439a0/BioinformaticsFMRP-TCGAbiolinksGUI-ec762c5" \ --library="C:/Users/asus/Documents/R/win-library/3.4" --install-tests

tiagochst commented 6 years ago

R has a limit of 100 DLL loaded. I believe that for R version 3.5 it has been solved. But, for the moment you will need to increase it manually. Please, just set the environment variable R_MAX_NUM_DLLS to R_MAX_NUM_DLLS=150 in the Renviron file.

There is some information in the vignettes. http://bioconductor.org/packages/release/bioc/vignettes/TCGAbiolinksGUI/inst/doc/index.html#21_installation

yinlisssss commented 6 years ago

i set the max number of DLL to 1000. it still exceeds.,,,, i wanna cry..

tiagochst commented 6 years ago

150 was sufficient. You need to restart R. Did you do that?

Tiago Chedraoui Silva

Em 23 de abr de 2018, à(s) 17:31, yinlisssss notifications@github.com escreveu:

i set the max number of DLL to 1000. it still exceeds.,,,, i wanna cry..

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

yinlisssss commented 6 years ago

1

yinlisssss commented 6 years ago

i am upating my R to 3.5,,,,let me see if it works later..

yinlisssss commented 6 years ago

finally , the installation of GUI 1.5.6succeeded, thank you very much !

but, ELMER analysis still doesnot work ...

Warning: Error in !: invalid argument type

tiagochst commented 6 years ago

I was not able to reproduce the error. Which ELMER version do you have ? 2.3.17? Also, which part does it stops?

yinlisssss commented 6 years ago

i checked my elmer, and it was an very old version.. stuipd i am...

now everything works well .

MUCH appreciated !!!