I just changed to a really big value.  (99999 was considered big many
years ago!)

On Thu, Jun 10, 2010 at 9:49 AM, Matthew Chambers
<[email protected]> wrote:
> Unless there are indeed downstream problems, it would be good to change that
> default upper scan number since others are running into the problem and it's
> a pretty obscure source of error.
>
> Dave, the peak filtering has been in msconvert for several months, so
> presumably users can now convert straight from MassHunter to MGF with
> msconvert. The remaining issue is if one wanted to convert to mzXML instead
> with both MS1 and MS2 scans in the file. If one only wanted to filter MS2s
> (like you certainly would want to do if you kept MS1s in profile) then
> that's not currently possible. It's a bit of a command-line syntax quagmire
> that we need to address in our tools.
>
> -Matt
>
> On 11/17/2009 12:52 PM, Brian Pratt wrote:
>>
>> Actually it will write 6 digit scan numbers as is.  The formatting just
>> specifies that it won't write any fewer than 5 digits.
>> I think the only code change I would want to make is to set that default
>> upper limit to the largest possible integer value, but as there's a
>> workaround I don't think we need to worry about it.
>> And yeah, who knows what happens downstream.  My intuition is that it
>> won't be a problem, actually, but that's only a guess.
>> Brian
>>
>> On Tue, Nov 17, 2009 at 8:44 AM, Dave Trudgian <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>>
>>    Brian,
>>
>>    I've just had a second look and found the reason the conversion
>>    stops at
>>    scan 99999:
>>
>>    The default upper scan number in the options struct is set to 99999:
>>
>>    MzXML2Search.cxx:117
>>    iLastScan = 99999;
>>
>>    Then just before the main loop it's used to set the value used to
>>    terminate the main loop:
>>
>>    MzXML2Search.cxx:344
>>     if (iAnalysisLastScan > options.iLastScan)
>>           iAnalysisLastScan = options.iLastScan;
>>
>>
>>    If the option default is changed to a higher value then the
>>    outputMGF(..) function would need to be changed to write >5 digit scan
>>    numbers. Don't know whether this would have an effect downstream :-(
>>
>>    If you would still like the mzML file as an example let me know
>>    where I
>>    can upload it. It's 1.5GB gzipped.
>>
>>    Cheers,
>>
>>    DT
>>
>>
>>    Brian Pratt wrote:
>>    > Returning to the behavior of mzXML2Search, just eyeballing the
>>    code I don't see any reason it should fail at scan numbers >
>>    99999.  Perhaps the problem is actually downstream from
>>    mzXML2Search?  While mzXML2Search should quite happily emit 6
>>    digit MGF scan numbers, I can imagine a consumer of MGF might not
>>    see that coming.
>>    >
>>    > Perhaps you could furnish an example of the msconvert output
>>    that's giving you trouble?  Eyeballing the code only gets one so
>>    far, it's ideal to see it actually running in the debugger.
>>    >
>>    > Brian Pratt
>>    >
>>    > On Tue, Nov 17, 2009 at 6:33 AM, Dave Trudgian
>>    <[email protected]
>>    <mailto:[email protected]><mailto:[email protected]
>>    <mailto:[email protected]>>> wrote:
>>    >
>>    > Matt,
>>    >
>>    > We just apply a (low) absolute threshold. Different values for
>>    different
>>    > instruments, but it's most critical on the Agilent and Water's
>>    QTOFs, as
>>    > without any threshold there are 1000s of peaks.
>>    >
>>    > Look forward to seeing the filtering in msconvert.
>>    >
>>    > DT
>>    >
>>    >
>>    > Matt Chambers wrote:
>>    >> What kind of thresholding do you do? Enabling that in msconvert is
>>    >> overdue - the backend code to support it is already in place.
>>    >>
>>    >> MzXML2Search is failing because it depends on a strict DTA name
>>    scheme
>>    >> with 5 digits. This is going to break with long LTQ Velos runs
>>    so it
>>    >> needs to be fixed regardless of the Agilent scanId issue. It's
>>    ironic
>>    >> that a Thermo file is the one to break the Thermo-centric
>>    assumptions. :P
>>    >>
>>    >> -Matt
>>    >>
>>    >>
>>    >> dctrud wrote:
>>    >>> MzXML2Search conversion to mgf fails for me on mzXML / mzML
>>    created
>>    >>> using msconvert from Agilent 6520 QTOF data.
>>    >>>
>>    >>> Trapper numbers spectra using an index starting at 1, whilst
>>    msconvert
>>    >>> uses the Agilent scan ID (can be a very large number).
>>    MzXML2Search
>>    >>> conversion of the resulting file into mgf then fails once it
>>    reaches a
>>    >>> scan ID > 99,999.
>>    >>>
>>    >>> Have seen similar problems with other programs and contacted Matt
>>    >>> Chambers who said that the numbering would stay the same, and that
>>    >>> it's better if programs which can't cope with the large
>>    numbers are
>>    >>> fixed. Is this possible (desirable?) for MzXML2Search and any
>>    other
>>    >>> TPP tools that might be affected?
>>    >>>
>>    >>> I can't use Trapper as I need to extract Profile MS + Centroid
>>    MS/MS
>>    >>> from a dual mode file (msconvert supported). I can't directly
>>    convert
>>    >>> to mgf with msconvert as I need to do peak thresholding to get
>>    file
>>    >>> sizes down to a reasonable level.
>>    >>>
>>    >>> Cheers,
>>    >>>
>>    >>> DT
>>    >>
>>    >
>>
>
> --
> You received this message because you are subscribed to the Google Groups
> "spctools-discuss" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to
> [email protected].
> For more options, visit this group at
> http://groups.google.com/group/spctools-discuss?hl=en.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"spctools-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/spctools-discuss?hl=en.

Reply via email to