Change Flag for unpaired reads from RNA seq

[Lien]

Dear all,

We performed an RNAseq experiment with 50bp single (unpaired) reads. The analysis was performed using CASAVA pipeline.

I would like to do some QC on the resulting bam-file. However, when I try using Picard (for example CollectRNASeqMetrics) it gives me the following error:
SAM validation error: ERROR: Record 1, Read name HWI-ST571_152:6:1204:18235:6093, First of pair flag should not be set for unpaired read.

I searched this forum, but I didn't really find out if and how I can change this FLAG so that Picard won't throw an error anymore and quits.

Many thanks
Lien

[maubp]

Have you reported this apparent bug in Casava?

Have you tried running Picard in lenient mode (so that it just warns about this FALG issue but doesn't treat it as an error)?

[Lien]

Oh, yes. How stupid, I forgot about the validation_stringency mode. The Picard-tool works now.

Thanks!