Whole exomes: Rsamtools, RAM issue

Question

[bioinfo8]

Hi,

I have many whole exome BAM files (aligned to reference). As an R admirer, I used Rstudio to do initial analysis on one of them (~2.7 GB) using Rsamtools.

1) Whether single or paired-end:

> testPairedEndBam("1.bam")   
[1] TRUE 
> quickBamFlagSummary("1.bam")           # Got detailed information

2) Read bam file

> bam <- BamFile("1.bam", asMates = TRUE)
> bam
    class: BamFile 
    path: 1.bam
    index:1.bam.bai
    isOpen: FALSE 
    yieldSize: NA 
    obeyQname: FALSE 
    asMates: TRUE                                       
    qnamePrefixEnd: NA 
    qnameSuffixStart: NA

3) Some high­ level information

> seqinfo(bam)

Seqinfo object with 1133 sequences from an unspecified genome:               # and other information

4) Read all the reads in the file using scanBam()

details <- scanBam(bam)

But at this step, it goes on running and running and I am stuck at this step. Any thoughts please? How much RAM is required to process a 3GB BAM file in Rstudio? I have windows 8.1, 64-bit computer with 16 GB RAM.

Thanks!

> sessionInfo()

R version 3.3.2 (2016-10-31)

Platform: x86_64-w64-mingw32/x64 (64-bit)

Running under: Windows >= 8 x64 (build 9200)

locale:

[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

attached base packages:

[1] stats graphics grDevices utils datasets methods base

loaded via a namespace (and not attached):

[1] zlibbioc_1.16.0 IRanges_2.4.8 XVector_0.10.0 futile.logger_1.4.3 parallel_3.3.2
[6] tools_3.3.2 GenomicRanges_1.22.4 lambda.r_1.1.9 futile.options_1.0.0 Biostrings_2.38.4
[11] S4Vectors_0.8.11 BiocGenerics_0.16.1 BiocParallel_1.4.3 Rsamtools_1.26.1 GenomeInfoDb_1.6.3
[16] stats4_3.3.2 bitops_1.0-6

[Benn]

Check in R how much memory you can use:

memory.limit()

[bioinfo8]

> memory.limit()
    [1] 16365

But if I wanted to check for a specific file, then?

[Benn]

I think memory is not the problem then. Do you get an error? How long do you let it run?

Maybe traceback can give you any clues?

traceback()

[bioinfo8]

It was running for 31 minutes and showed following error:

> details <- scanBam(bam) 
Error in value[[3L]](cond) : cannot allocate vector of size 1024.0 Mb
  file: 1.bam
  index: 1.bam.bai

> traceback()
8: stop(conditionMessage(err), "\n  file: ", path(file), "\n  index: ", 
       index(file))
7: value[[3L]](cond)
6: tryCatchOne(expr, names, parentenv, handlers[[1L]])
5: tryCatchList(expr, classes, parentenv, handlers)
4: tryCatch({
       .Call(func, .extptr(file), space, flag, simpleCigar, tagFilter, 
           mapqFilter, ...)
   }, error = function(err) {
       stop(conditionMessage(err), "\n  file: ", path(file), "\n  index: ", 
           index(file))
   })
3: .io_bam(.scan_bamfile, file, reverseComplement, yieldSize(file), 
       tmpl, obeyQname(file), asMates(file), qnamePrefix, qnameSuffix, 
       param = param)
2: scanBam(bam)
1: scanBam(bam)

[Benn]

Since you are using windows, maybe this answer on stackoverflow can help:

http://stackoverflow.com/questions/5171593/r-memory-management-cannot-allocate-vector-of-size-n-mb/24754706#24754706

[bioinfo8]

Thanks! but did not find it useful.

[RJ]

Having the same issue in 2020. Rsamtools's scanBam will overflow the RAM and whole PC freezes. I'm on Ubuntu 20.04.