.BAM file wonky after SMRA?

[kga1978]

Sorry, title is wrong - make that SRMA...

I am trying to perform local realignment on a couple of BAM alignment files. I got stuck with GATK and tried SRMA - seems more intuitive and easier to use. However, the output BAM files I have gotten appear to be 'broken' - I think because of quality scores?

Here's what I did:
java -Xmx2048m -jar /usr/local/bin/srma/srma-0.1.15.jar I=G502sorted.bam O=G502realigned.bam R=LASV096/Mosaik/reference.fasta

samtools index G502realigned.bam

The programs complete successfully, but the resultant .bam file can't be read by FastQC and when I try to import it into Geneious I get a 'phred score above 102' error. Did SMRA somehow change my quality scores? Any way I can correct for this? I can open up the file in Tablet and it has been realigned, but I can see quality scores, etc. in Tablet.

Thanks

[kga1978]

I also found another funky thing. In one of the regions where the alignment was refined, some of the reads before the 'break point' are now no longer aligned correctly. Here's an image from IGV (realigned on top):



The incorrectly aligned read to the right of the SNP should have been realigned and the two reads to the left shouldn't have been touched at all. Any thoughts as to how I can control this?

Still trying to get the first issue resolved as well...

[nilshomer]

If you could give a small test case for the quality out of range I can debug.

I assume you are re-aligning SOLiD reads? If you could extract the reads in that region I can try to debug that too.

[kga1978]

Hi nils,

Sorry, I'm aligning illumina reads, not SOLiD. The BAM files are very small so I'm just going to attach the files:

http://cl.ly/42261j1z1L3d3R022S0A

Any thoughts on the realignment issue?

Let me know if you need anything else.

[nilshomer]

Could you also send the reference? I don't have that one...

[kga1978]

Yes - sorry.

http://cl.ly/0Q2Z2a1U3p403u0N3g1f