Using RIPSeeker : RIPSeeker terminates with the "cannot allocate memory" error

Question

[rew_27]

Hello all,

I am analyzing RIPSeq data and using the RIPSeeker package for it. I have a total of 7 samples with 3 biological replicates each. The three replicates correspond to: an input sample, an IgG negative control and the knockdown sample. Thus, I have a total of (7*3 ) 21 samples.

I am analyzing one triplicate at a time. I would first like to know if this is the right way to go. My script is as follows:

source("http://bioconductor.org/biocLite.R")

biocLite("RIPSeeker")
library(RIPSeeker)
library(GenomicRanges)

biomart <- "ensembl"
biomaRt_dataset <- "hsapiens_gene_ensembl"
cName = "Ctrl"

bamfiles = list.files(path = "/Users/rewati/bam", recursive=TRUE, full.names=TRUE)
seekOut = ripSeek(bamPath = bamfiles, cNAME = cName, binSize = NULL, minBinSize = 10000, maxBinSize = 12000, paired = TRUE, reverseComplement = TRUE, genomeBuild = "hg19", uniqueHit = TRUE, biomaRt_dataset = biomaRt_dataset, multicore=TRUE, exportFormat = GFF, pvalCutoff = 0.05, pvalAdjCutoff = 0.05, eFDRCutoff = 0.05, outDir="/Users/rewati/ripseekout")

After running the script, the script halts with this error message:

Error in mcfork() : 
  unable to fork, possible reason: Cannot allocate memory
Calls: ripSeek -> mainSeek -> mclapply -> lapply -> FUN -> mcfork
Execution halted
Warning message:
system call failed: Cannot allocate memory

I then tried the script with the multicore option set to FALSE but then it takes forever to give results. Any insights on how to use the package and how to speed up the analysis will be appreciated!

[jared.andrews07]

The error is pretty straightforward - you don't have the memory to run it with all processing cores available. Either run it on a cluster with the appropriate resources or don't use the multicore option - those are pretty much your only choices.