coverageBED - Killed

[Thai Cao]

I am trying to used coverageBED using the following input files:
- BAM file of paired-end reads aligned with Novoalign
- BED file of genomic windows from Agilent tiling the exon coverage
from their SureSelect XT Mouse Exon kit.

This is the output I get:
medline@medline:~/8060R/test$ coverageBed -abam
8060X2_110616_SN141_0362_BB02JRABXX_6.mate.dup.sort.sort.bam -b
027612_D_BED_20110119.bed > result_8060X2.txt
Killed
medline@medline:~/8060R/test$

I am using a desktop running uBuntu 11.04 with 4GB memory. This is an
upgrade from 2GB memory where the system would just crash without
getting the killed error message. Am I getting the killed errors
because of inadequate system memory?

[Aaron Quinlan]

Hi,

Yes, memory consumption may be the issue here. One thing you can do (besides use a machine with more memory) is use the -counts option, which uses much less memory than the default option.

Best,
Aaron

[Thai Cao]

Thank you for your reply and you wonderful BEDTools application suite.

Interestingly, intersectBED just completed overnight with the same
input files no problem:

intersectBed -abam
8060X1_110616_SN141_0362_BB02JRABXX_5.mate.dup.sort.sort.bam -b
027612_D_BED_20110119.bed -bed

System memory cost is relatively low right now, easy for me to
acquire, may I ask what users typically configure for BEDTools, 8GB?
16GB or higher?

On Feb 9, 4:08 am, Aaron Quinlan <aaronquin...@gmail.com> wrote:
> Hi,
>
> Yes, memory consumption may be the issue here.  One thing you can do (besides use a machine with more memory) is use the -counts option, which uses much less memory than the default option.
>
> Best,
> Aaron
>
> On Feb 8, 2012, at 5:49 PM, Thai Cao wrote:
>
>
>
>
>
>
>
> > I am trying to used coverageBED using the following input files:
> > - BAM file of paired-end reads aligned with Novoalign
> > - BED file of genomic windows from Agilent tiling the exon coverage
> > from their SureSelect XT Mouse Exon kit.
>

> > This is the outpintersectBed -abam 8060X1_110616_SN141_0362_BB02JRABXX_5.mate.dup.sort.sort.bam -b 027612_D_BED_20110119.bed -bedut I get:

[Thai Cao]

Well, actually using the -count option like so just completed without
error:

coverageBed -count -abam
8060X1_110616_SN141_0362_BB02JRABXX_6.mate.dup.sort.sort.bam -b
027612_D_BED_20110119.bed > result_8060X1.txt

Amazingly, the run only took a little over 1 hour on my vanilla AMD
Athlon II X2 2.8GHZ with 4GB DDR2 box running uBuntu 11.04. The input
BAM file had 100x10^6 paired-end aligned reads (5GB file) and the BED
file had 5x10^5 windows.

Many thanks.

On Feb 9, 4:08 am, Aaron Quinlan <aaronquin...@gmail.com> wrote: