Re: [proteowizard-developer] pwiz reading terribly slow for Xcalibur exported mzML

[Matthew C. <mat...@gm...>]

There is a validator: http://www-bs2.informatik.uni-tuebingen.de/services/OpenMS/mzML/

Who shall we hire to hold a gun to their head to use it? :) Jim Shoftahl told me that this Bioworks 
exporter is dead code. He has a standalone converter and Proteome Discoverer 1.3 has one as well. 
Perhaps those are better (and tested).

-Matt


On 7/21/2011 10:49 AM, Brendan MacLean wrote:
> Once again, a problem with having a "standard" format without a validation test suite to ensure
> people who implement the format get it right.  Unless you are going to have only one implementation,
> you really need a validation test suite to guarantee consistency and correctness.
>
> On Thu, Jul 21, 2011 at 4:12 AM, Steffen Neumann <sne...@ip...
> <mailto:sne...@ip...>> wrote:
>
>     Hi,
>
>     On Wed, 2011-07-20 at 16:50 +0200, Steffen Neumann wrote:
>      > No, that's not the comparison. I have a *.RAW
>      > that was converted to mzML via 1) Xcalibur and 2) msconvert.
>
>     So, just for the archives, it seems that the Xcalibur-created mzML
>     is at least a bit strange. As you see, all spectra have index="1",
>     which caused pwiz to re-index the whole file for each of the spectra.
>
>     <spectrum index="1" id="controllerType=MS controllerNumber=1 scan=1" defaultArrayLength="1517">
>     <spectrum index="1" id="controllerType=MS controllerNumber=1 scan=2" defaultArrayLength="97">
>     <spectrum index="1" id="controllerType=MS controllerNumber=1 scan=3" defaultArrayLength="1588">
>             ...
>     <spectrum index="1" id="controllerType=MS controllerNumber=1 scan=1600" defaultArrayLength="110">
>     <spectrum index="1" id="controllerType=MS controllerNumber=1 scan=1601" defaultArrayLength="1033">
>     <spectrum index="1" id="controllerType=MS controllerNumber=1 scan=1602" defaultArrayLength="92">
>
>     Matt knows about this, and a fix/workaround is in the works.
>
>     Yours,
>     Steffen