Novoalign inappropriately softclipping?

[myrna]

I have downloaded some exome data that appears to have been aligned using Novoalign
I determined this from the header:

@PG ID:novoalign VN:V2.06.09 CL:novoalign -c 16 -o SAM -o SoftClip -r Random -f 1.fastq 2.fastq -d /home/vg37/genomes/hg18_nohaps.fa.novocraft

When looking at the data in IGV, I see that lots of reads have a large fraction of the sequence softclipped, resulting in cigar strings such as "7S36M58S". The image below shows a region of the genome where this is particularly bad. I noticed that the softclipped sequences often match the genome sequence. I'm wondering why these regions were softclipped in the first place. Is this a known bug?

Ryan

[sparks]

Softclipping

Hi,

I've been back through the change logs and can't find any record of fixing a bug that might cause this. Large amounts of soft clipping like this could be caused by low base qualities especially if quality is 2 (#) as we treat this almost like an N. In browser were you shading bases by quality?

Colin

[sparks]

BAM Extract

Could you send be a extract (samtools view) from the bam file for the region you displayed and I'll take a look at it. Email to support at novocraft-dot-com.
Colin