"sambamba-view: not enough data in stream"

[scatreux]

Hi,

I'm getting an error: "sambamba-view: not enough data in stream" by running this command line:
/usr/local/bin/sambamba view -F 'mapping_quality > 0' -L /exome-12-callable-regions.bed -f bam -l 1 /bamfile1.bam > test.bam

I have done the following tests:

1. removing the -L BED file from the command line gets rid of the error
2. swapping the bamfile1.bam for a different bamfile2.bam and keeping the -L BED file in the command line also gets rid of the error.

So, it seems it is when the -L BED file and bamfile1.bam are in the same command line that I get the error. Can you please provide tips on how I can further narrow down on the issue?
I'm also happy uploading the BED and bamfile1.bam to a suggested place if it is helpful for debug.
I tried a samtools view of the -L BED and bamfile1.bam on the same command line and
that also completed successfully.

Many thanks
Severine

[lomereiter]

Hi,

Did you try the trunk version?(https://github.com/lomereiter/sambamba/releases/download/latest/sambamba_latest_centos5.tar.bz2)

[scatreux]

Hi
I did. I still get the error.

[lomereiter]

OK, could you upload the files somewhere, dropbox or the like?

[scatreux]

Hi Brad

Is it possible to share the location
Of the bam and bed files I uploaded
Last night?

Tx
Severine

Sent from my iPhone

On Jun 19, 2014, at 6:03 AM, Artem Tarasov notifications@github.com wrote:

OK, could you upload the files somewhere, dropbox or the like?

—
Reply to this email directly or view it on GitHub.

[scatreux]

I can’t use Dropbox because the BAM file is 6.9GB.

Do you have an FTP site I can upload the files to?

Tx!

Severine

From: Artem Tarasov [mailto:notifications@github.com]
Sent: Thursday, June 19, 2014 6:03 AM
To: lomereiter/sambamba
Cc: scatreux
Subject: Re: [sambamba] "sambamba-view: not enough data in stream" (#79)

OK, could you upload the files somewhere, dropbox or the like?

—
Reply to this email directly or view it on GitHub #79 (comment) . https://github.com/notifications/beacon/6052383__eyJzY29wZSI6Ik5ld3NpZXM6QmVhY29uIiwiZXhwaXJlcyI6MTcxODgwMjE4MiwiZGF0YSI6eyJpZCI6MzUwMTc3NjB9fQ==--ab1f869fc7b1373fc1276e182d7ae31db2d0577c.gif

[chapmanb]

Artem and Severine;
Here are the problem BAM and BED files. Severine, thanks for uploading them. Artem, sorry I haven't had a chance to reduce this for something more minimal but hope it helps as is:

https://s3.amazonaws.com/bcbio_nextgen_upload/NA12878_NGv3_exome_GRCh37-dedup.bam
https://s3.amazonaws.com/bcbio_nextgen_upload/NA12878_NGv3_exome_GRCh37-dedup-12-callable-regions.bed

[lomereiter]

Thanks Brad, I'm downloading them now.
For future reference, http://disk.yandex.com provides a cloud storage service with 10GB for free.

[scatreux]

many thanks Brad and Artem. just checking if anything was revealed from these files?

[lomereiter]

Not yet, I've just started to minimize the case.

[lomereiter]

Reduced it down to 30KB: https://dl.dropboxusercontent.com/u/7916095/12_4.bam, will now investigate.

[scatreux]

Many thanks

From: Artem Tarasov [mailto:notifications@github.com]
Sent: Thursday, June 19, 2014 10:32 PM
To: lomereiter/sambamba
Cc: scatreux
Subject: Re: [sambamba] "sambamba-view: not enough data in stream" (#79)

Reduced it down to 30KB: https://dl.dropboxusercontent.com/u/7916095/12_4.bam, will now investigate.

—
Reply to this email directly or view it on GitHub #79 (comment) . https://github.com/notifications/beacon/6052383__eyJzY29wZSI6Ik5ld3NpZXM6QmVhY29uIiwiZXhwaXJlcyI6MTcxODg2MTUxMCwiZGF0YSI6eyJpZCI6MzUwMTc3NjB9fQ==--82e45897d4259c132b7463d12eb1e4d517dc4c77.gif

[lomereiter]

The issue was integer overflow in one of routines. The issue manifests itself when one of BGZF blocks is fully filled (65kB).
Thanks for testing!

[scatreux]

Great! many thanks.

Brad,

Is it possible to integrate latest sambamba view to bcbio-nextgen?

Severine

From: Artem Tarasov [mailto:notifications@github.com]
Sent: Friday, June 20, 2014 5:32 AM
To: lomereiter/sambamba
Cc: scatreux
Subject: Re: [sambamba] "sambamba-view: not enough data in stream" (#79)

The issue was integer overflow in one of routines. The issue manifests itself when one of BGZF blocks is fully filled (65kB).
Thanks for testing!

—
Reply to this email directly or view it on GitHub #79 (comment) . https://github.com/notifications/beacon/6052383__eyJzY29wZSI6Ik5ld3NpZXM6QmVhY29uIiwiZXhwaXJlcyI6MTcxODg4NjczNSwiZGF0YSI6eyJpZCI6MzUwMTc3NjB9fQ==--28be6a690200a2bde5e8e7528e7d81f816a8ba98.gif