Thanks Tim and Nils. Thinking about it some more after your responses, I
agree it is worth checking this situation. If 0x1 is unset then there is no
need for the bam author to set the other bits like first in pair or last in
pair (assuming here that by default all flags are off and setting them is
an active decision by the bam author). The fact that they did set these
flags is suspicious. I will contact the bam author to let them know the
preference on this issue.
--
Eric Roller
On Fri, Sep 12, 2014 at 5:22 AM, Nils Homer <[email protected]>
wrote:
> Hey Eric,
>
> you can set the VALIDATION_STRINGENCY to SILENT or LENIENT to remove or
> reduce the number of checks. In some cases, while the read may be allowed
> under the SAM specification, the requirement is essential for the execution
> of the program.
>
> I would believe that having the first of pair flag set for an unpaired
> reads should only be done in exceptional circumstances, and the data not
> shared with others in such a form.
>
> Sincerely,
>
> Nils Homer
>
> On Fri, Sep 12, 2014 at 3:22 AM, Tim Fennell <[email protected]>
wrote:
> Hi Eric,
>
> There has been a long-running difference of opinion on topics like this,
> with one group pushing the view that is in the spec (i.e. behaviour like
> this should be left unspecified) and there’s another group (including me)
> that believes that SAM & BAM are complex enough formats and so many
> mistakes are made creating BAM files that it is better to test for
> situations like this as they most likely indicate an error, perhaps one of
> consequence, somewhere earlier in the processing chain.
>
> If you would prefer to ignore this error with ValidateSamFile you can
> selectively ignore validations you find too strict with the IGNORE= option.
>
> -t
>
On Fri, Sep 12, 2014 at 2:22 AM, Eric Roller <[email protected]> wrote:
>
>> Using version 1.119 I get this error:
>>
>> First of pair flag should not be set for unpaired read
>>
>>
>> Basically, Picard is complaining that 0x40 is set when 0x1 is unset.
>> However, from the samtools specification:
>>
>> "If 0x1 is unset, no assumptions can be made about 0x2, 0x8, 0x20, 0x40 and
>> 0x80."
>>
>>
>> Furthermore, the 0x40 description from the spec says
>>
>> 0x40 the first segment in the template
>>
>> For an unpaired read it will be both the first and last segment in the
>> template. That doesn't mean it is paired so there should be no error here.
>>
>>
>> Am I missing something?
>>
>>
>> Thanks,
>>
>> --
>> Eric Roller
>>
>>
>> ------------------------------------------------------------------------------
>> Want excitement?
>> Manually upgrade your production database.
>> When you want reliability, choose Perforce
>> Perforce version control. Predictably reliable.
>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Samtools-help mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/samtools-help
>>
>>
>
------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Samtools-help mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/samtools-help
