BWA PE Problem

[puggie]

Ok thanks for your replies. I tried with new fastq files, and still got the problem, only one sample passed through to the complete SAM file. I also tried mapping without any preprocessing, and same error BWA stops at sampe.

I want to run it in command line outside of Galaxy, but Im not super-familiar with linux. Right now Im indexing the genome. Should all indexing files be in the same directory as my input fastqs and reference? Or can I specify the location somehow? I just installed bwa by apt-get on a clean Ubuntu disk image. But where exactly does it install? Only thing I know, is that bwa is in the PATH.

And my apologies for such basic linux questions.

[Kennels]

As mentioned above, you need to make sure that your processed paired end reads are paired correctly in read 1 and read 2 files. I had BWA stall forever when my reads weren't paired properly after preprocessing (filtering by quality will get rid of different reads in read 1 and read 2 file if you don't use the appropriate software or re-pair after processing).

[swbarnes2]

What does bwa sample tell you are the approximate insert sizes? If they are crazy huge, it can take the software a long time to decide that, and move on.

But no, it should not hang. It should be outputting its progress every couple of seconds.

[puggie]

With the latter scenario I would have expected BWA to "hang" for several hours/days? Im currently running an alignment with newly transferred fastqs, and will report back if that doesnt solve the issue.

[swbarnes2]

Pasting the output when it goes bad might help.

Is it possible that the fastqs got corrupted part-way through?

Renaming and trimming shouldn't hurt things, as long as the millionth read in fastq 1 really is the mate of the millionth read of fastq 2, and so on.

[puggie]

Dear forum,

I am experiencing problems during BWA paired-end mapping on my local setup. I have two fastq files for each sample. BWA aln step seems okay, but then at sampe the alignment may fail after outputting 1 Mb or even 1 Gb of SAM data. Could this happen to be a pre-processing issue? Fastq --> rename --> trimming ?

BWA is executed under Galaxy platform and I dont see the log file, so I may try to run it on command line. I see it fails at sampe via htop command.

I have also indexed the reference.