"missing colon in auxiliary data" error when working with fastq's

[Peter Waltman]

Are there any switches that allow for intermediate files to be generated while running SMuFin?

I'm getting an error regarding "missing colon in auxiliary data" that I've traced to bam_import.c (complete error below). Based on discussions I've found online, it appears the error has something to do with missing information in one of the fields after the reads have been aligned, e.g. this and this.  However, with no intermediate file to re-examine, it is impossible to determine what read is causing samtools to throw this error.

Parse error at line 15165: missing colon in auxiliary data

[juggernaut:30570] *** Process received signal ***

[juggernaut:30570] Signal: Aborted (6)

[juggernaut:30570] Signal code:  (-6)

[juggernaut:30570] [ 0] /lib64/libpthread.so.0() [0x3a45a0f710]

[juggernaut:30570] [ 1] /lib64/libc.so.6(gsignal+0x35) [0x3a45232625]

[juggernaut:30570] [ 2] /lib64/libc.so.6(abort+0x175) [0x3a45233e05]

[juggernaut:30570] [ 3] SMuFin() [0x53d78c]

[juggernaut:30570] [ 4] SMuFin(sam_read1+0xf4d) [0x53ef40]

[juggernaut:30570] [ 5] SMuFin(samread+0x70) [0x52a0ce]

[juggernaut:30570] [ 6] SMuFin(main_samview+0x4b4) [0x5288c5]

[juggernaut:30570] [ 7] SMuFin(gsuffix_co_asm_impl+0x25) [0x48813d]

[juggernaut:30570] [ 8] SMuFin(gsuffix_co+0x2b) [0x47f30a]

[juggernaut:30570] [ 9] SMuFin(_Z18process_block_funcPv+0x98) [0x466bb8]

[juggernaut:30570] [10] /lib64/libpthread.so.0() [0x3a45a079d1]

[juggernaut:30570] [11] /lib64/libc.so.6(clone+0x6d) [0x3a452e88fd]

[juggernaut:30570] *** End of error message ***

[SMufin]

Dear Peter,

The smufin version of the webpage does not include an option that allows intermetiate files to be generated while running the application. However, as you mention, it seems the problem is related to the format of some field after reads have been aligned, which is related to samtools.

One thing you may try is to reproduce the error by using independently samtools, for example: first, converting from fastq to sam and then to bam.
It is something like this post mentions http://sourceforge.net/p/bio-bwa/mailman/message/31803286/.

As it is shown in that post, if you reproduce this problem, the samtool can give you the line in you input file (and the read) that generates the problem.

Best regards,
Smufin team.

[Peter Waltman]

Ok, well, I can easily generate a sam file since I have bam files for this data, but things that I need to know are:

a) how are you converting the fastq's to sam? are you doing any sort of alignment? If so, what method are you using? BWA v. Bowtie? Something else?

b) if you are aligning with BWA, are you aligned with BWA-MEM or BWA-ALN?

b.2) if you are using BWA-MEM, what flags should I specify?

c) should the final sam/bam file be name-sorted or coordinate sorted?

d) should the final sam/bam file be de-dup'd?

Thanks!