Biostar Beta. Not for public use.
ChIPQC showed errors 'nnodes' must be >= 1
0
Entering edit mode
5.2 years ago
lkw.wang • 0
United States

When I used ChIPQC, I followed the manuals to construct a sample sheet (name is "examplefile.csv"), and then run:

b = aa = ChIPQC(a, annotation=NULL)

air_k14_2 air_k14_2 air1 1 bed
air_k14_8 air_k14_8 air1 2 bed
air_k23_3 air_k23_3 air1 1 bed
air_k23_9 air_k23_9 air1 2 bed
air_k9_25 air_k9_25 air1 1 bed
air_k9_27 air_k9_27 air1 2 bed
C2H4_k14_5 C2H4_k14_5 C2H4-1 1 bed
C2H4_k14_11 C2H4_k14_11 C2H4-1 2 bed
C2H4_k23_6 C2H4_k23_6 C2H4-1 1 bed
C2H4_k23_12 C2H4_k23_12 C2H4-1 2 bed
C2H4_k9_26 C2H4_k9_26 C2H4-1 1 bed
C2H4_k9_28 C2H4_k9_28 C2H4-1 2 bed
Computing metrics for 16 samples...
Error in (function (nnodes = getOption( "mc.cores", 2L), ...) :
'nnodes' must be >= 1

When I use the same "examplefile.csv" file for DiffBind analysis, it works very well.

So, I don't know why the ChIPQC doesn't work.

Any help would be appreciated!!

0
Entering edit mode

Looks like it might have to do with you running ChIPQC in a non-multicore environment. You could try re-running and specifying more cores. Also, for bioconductor related questions, you would get a quicker response here: https://support.bioconductor.org/ (make sure you tag your question the the package that you are having problems with)

0
Entering edit mode

Thank you very much for your reply. I used the detectCores function to determine I am using a multicore environment. I will refer to the forum you suggested though. Thanks again.

detectCores()
[1] 8
detectCores(logical = FALSE)
[1] 4

0
Entering edit mode

Thanks for your answer! I will learn how to define the cpu numbers in R.

0
Entering edit mode
19 months ago
Netherlands

As Ying said and I speculate the same, either the number of CPUs are not defined before running the above function or the value is not increase from 1 (to have it a parallel computing envrionment)

Check the function here http://code.metager.de/source/xref/gnu/R/src/library/parallel/R/unix/forkCluster.R

parallel R package is throwing that error.

Check how to allocate/assign CPUs to R before running the program. detectCores() just tells you how many you have

Check snowpackage and makeCluster function.

This will get you started

https://wiki.hpcc.msu.edu/display/hpccdocs/R

https://cran.r-project.org/web/views/HighPerformanceComputing.html

http://dirk.eddelbuettel.com/papers/useR2010hpcTutorial.pdf

HTH