Hi Alex,
Yeah, I assumed something like that. I have also frequently run into
annoyances related to BAMs produced elsewhere and picard's requirements.
Would it make sense to have either, 1) a standalone CleanBam tool or
better CleanSam tool that handles BAMs, or 2) some more explicit arguments
to MergeBamAlignment to make it cleanup a single BAM more explicitly?
-Ben
On Thu, Jul 31, 2014 at 9:56 AM, Alec Wysoker <al...@broadinstitute.org>
wrote:
> Hi Ben,
>
> You are right that we should document this better. And, no, I don't mean
> CleanSam.
>
> Various aligners do all sorts of strange things to the reads; e.g. not all
> the reads are written to the output, reads are hard-clipped, some
> attributes in the input don't make it to the output, etc.
> MergeBamAlignment takes as input 2 BAMs: an unmapped BAM that presumably
> satisfies Picard's validation requirements, and an aligned BAM. It takes
> only the alignment information from the aligned BAM, and everything else
> from the unmapped BAM, and produces an aligned BAM that satisfies strict
> Picard validation.
>
> -Alec
>
>
> On 7/31/14, 12:25 PM, Ben Bimber wrote:
>
> Hi Alex,
>
> Using MergeBamAlignment as a cleanup tool (what you seem to be
> suggesting) is not something I would have guessed or seen in the docs. Did
> you mean CleanSam or is merge doing something extra beyond CleanSam? Is
> it a workaround to let you work in BAM versus SAM?
>
> -Ben
>
>
> On Thu, Jul 31, 2014 at 9:15 AM, Alec Wysoker <al...@broadinstitute.org>
> wrote:
>
>> Hi Wolfgang,
>>
>> I will not argue that Picard's behavior in this regard is absolutely
>> conforming to the spec. If you troll through the list archive I believe
>> you will find that this had been discussed before. We have found this
>> level of validation to be very useful. If these alignments are coming
>> from BWA, we recommend using MergeBamAlignment which will create a BAM
>> that satisfies Picard's admittedly very strict validation. Or, you can
>> disable the validation.
>>
>> -Alec
>>
>> On 7/31/14, 11:40 AM, Wolfgang Maier wrote:
>> > On 31.07.2014 17 <31.07.2014%2017>:26, Alec Wysoker wrote:
>> >> See the Picard FAQ:
>> >>
>> >>
>> http://sourceforge.net/p/picard/wiki/Main_Page/#q-why-am-i-getting-errors-from-picard-like-mapq-should-be-0-for-unmapped-read-or-cigar-should-have-zero-elements-for-unmapped-read
>> >>
>> >> -Alec
>> >>
>> > Well, that shows that it is known behavior of PicardTools, but still, is
>> > it *correct* behavior given the format specifications ?
>> >
>> >> On 7/31/14, 4:36 AM, Wolfgang Maier wrote:
>> >>> Hi,
>> >>>
>> >>> I've tried to run SamToFastq on a sam file with unaligned reads. Right
>> >>> on the first line picardtools complains:
>> >>>
>> >>> Exception in thread "main"
>> >>> htsjdk.samtools.SAMFormatException: Error parsing text SAM
>> >>> file. MAPQ must be zero if RNAME is not specified; Line 4
>> >>> Line:
>> >>> HWI-EA332_1_2_662_117/1 4 * 0 255 * *
>> 0 0
>> >>> CCACACTACACCCTAACACACCCTAATCTAACCCTGGCCAACCTGTCTCTC
>> >>> BBBBCCCCBB@BCB?CBC@ABB?<>AB?CBCBB@BC>BB@BB7BCB<BCA? RG:Z:007
>> >>>
>> >>> After consulting the sam format specifications I think that this
>> >>> behavior is incorrect.
>> >>> The read has bit 4 set in the FLAG field and with that "no assumption
>> >>> can be made about RNAME, POS, CIGAR, MAPQ", so picard shouldn't even
>> >>> look at MAPQ here.
>> >>> Even if it did, I don't know whether it should expect MAPQ to be zero
>> in
>> >>> the absence of RNAME as a MAPQ value of 255 "indicates that the
>> mapping
>> >>> quality is not available", which seems to be about as reasonable as
>> >>> setting it to 0.
>> >>>
>> >>> Please correct me if I'm getting this wrong,
>> >>> Wolfgang
>> >>>
>> >>>
>> >>>
>> ------------------------------------------------------------------------------
>> >>> Infragistics Professional
>> >>> Build stunning WinForms apps today!
>> >>> Reboot your WinForms applications with our WinForms controls.
>> >>> Build a bridge from your legacy apps to the future.
>> >>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
>> >>> _______________________________________________
>> >>> Samtools-help mailing list
>> >>> Samtools-help@lists.sourceforge.net
>> >>> https://lists.sourceforge.net/lists/listinfo/samtools-help
>> >>
>> >>
>> ------------------------------------------------------------------------------
>> >> Infragistics Professional
>> >> Build stunning WinForms apps today!
>> >> Reboot your WinForms applications with our WinForms controls.
>> >> Build a bridge from your legacy apps to the future.
>> >>
>> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
>> >>
>> >
>> ------------------------------------------------------------------------------
>> > Infragistics Professional
>> > Build stunning WinForms apps today!
>> > Reboot your WinForms applications with our WinForms controls.
>> > Build a bridge from your legacy apps to the future.
>> >
>> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
>> > _______________________________________________
>> > Samtools-help mailing list
>> > Samtools-help@lists.sourceforge.net
>> > https://lists.sourceforge.net/lists/listinfo/samtools-help
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Infragistics Professional
>> Build stunning WinForms apps today!
>> Reboot your WinForms applications with our WinForms controls.
>> Build a bridge from your legacy apps to the future.
>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Samtools-help mailing list
>> Samtools-help@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/samtools-help
>>
>
>
>
------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls.
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Samtools-help mailing list
Samtools-help@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/samtools-help