psidev-ms-dev

[Psidev-ms-dev] dataProcessing

[Marc S. <st...@in...>]

Hi all,

i have problems understanding how the data processing part of the schema 
is intended.
The main question I have is: what does the "order" attribute define?
Is it the order of different tools that were applied or the order of 
several processing steps done by the same software?

As it is under "processingMethod", I thought it defines the order of 
several processing steps done by the same software.
However the terms "data processing action" and "file format conversion" 
should not be repeatable in this case.
Otherwise the order is undefined again.

When i looked into the tiny1 example, it confused me even more.

I attached an image that shows parts of the corresponding parts of the 
example file, the CV file, the schema file and the mapping file.

Best,
  Marc

Re: [Psidev-ms-dev] dataProcessing

[Eric D. <ede...@sy...>]

Hi Marc, below:

> From: psi...@li...
[mailto:psidev-ms-dev-
> 
> Hi all,
> 
> i have problems understanding how the data processing part of the
schema
> is intended.
> The main question I have is: what does the "order" attribute define?
> Is it the order of different tools that were applied or the order of
> several processing steps done by the same software?

It is the order of processing steps applied to the dataset over its
journey from mass spec to current form.

> As it is under "processingMethod", I thought it defines the order of
> several processing steps done by the same software.
> However the terms "data processing action" and "file format
conversion"
> should not be repeatable in this case.
> Otherwise the order is undefined again.

I don't understand what you mean by this. It would seem entirely
possible for vendor software A to perform some thresholding first and
write out data in thresholded profile mode. Then perhaps FOSS software B
might be used to convert to mzML. Then some other program might be used
to centroid the data and write out another mzML file. These might be 3
softwares used in the history of the data with order 1, 2, 3,
respectively.

> When i looked into the tiny1 example, it confused me even more.

In this case there are several algorithms applied in the same step since
it is not known precisely in what order they were performed. So in the
example, first Xcalibur was used to perform deisotoping, charge
deconvolution, and peak picking. These are all lumped together as step
1. Perhaps they are result of a single algorithm or action and not
separatable. Or perhaps there should be some inherent order, but it is
not know to the writer. Then step 2 is the conversion to mzML. If the
order of deisotoping, charge deconvolution, and peak picking is
specifically known and relevant, it would be permissible to write this
is a 4-step process, with order 1, 2, 3, and 4.

Has this answered your question? Or perhaps I have misunderstood your
confusion? If this clears it up, I will update the documentation to
reflect this description.

Also, if any of the other designers feel I have not described the intent
correctly, please speak up!

Regards,
Eric



> I attached an image that shows parts of the corresponding parts of the
> example file, the CV file, the schema file and the mapping file.
> 
> Best,
>   Marc

Re: [Psidev-ms-dev] dataProcessing

[Marc S. <stu...@gm...>]

Hi Eric,

thanks for your reply.
When you want to allow both ordered and unordered processing steps, it 
all makes sense.
I thought everything should be ordered.

Regards,
  Marc

Eric Deutsch wrote:
> Hi Marc, below:
>
>   
>> From: psi...@li...
>>     
> [mailto:psidev-ms-dev-
>   
>> Hi all,
>>
>> i have problems understanding how the data processing part of the
>>     
> schema
>   
>> is intended.
>> The main question I have is: what does the "order" attribute define?
>> Is it the order of different tools that were applied or the order of
>> several processing steps done by the same software?
>>     
>
> It is the order of processing steps applied to the dataset over its
> journey from mass spec to current form.
>
>   
>> As it is under "processingMethod", I thought it defines the order of
>> several processing steps done by the same software.
>> However the terms "data processing action" and "file format
>>     
> conversion"
>   
>> should not be repeatable in this case.
>> Otherwise the order is undefined again.
>>     
>
> I don't understand what you mean by this. It would seem entirely
> possible for vendor software A to perform some thresholding first and
> write out data in thresholded profile mode. Then perhaps FOSS software B
> might be used to convert to mzML. Then some other program might be used
> to centroid the data and write out another mzML file. These might be 3
> softwares used in the history of the data with order 1, 2, 3,
> respectively.
>
>   
>> When i looked into the tiny1 example, it confused me even more.
>>     
>
> In this case there are several algorithms applied in the same step since
> it is not known precisely in what order they were performed. So in the
> example, first Xcalibur was used to perform deisotoping, charge
> deconvolution, and peak picking. These are all lumped together as step
> 1. Perhaps they are result of a single algorithm or action and not
> separatable. Or perhaps there should be some inherent order, but it is
> not know to the writer. Then step 2 is the conversion to mzML. If the
> order of deisotoping, charge deconvolution, and peak picking is
> specifically known and relevant, it would be permissible to write this
> is a 4-step process, with order 1, 2, 3, and 4.
>
> Has this answered your question? Or perhaps I have misunderstood your
> confusion? If this clears it up, I will update the documentation to
> reflect this description.
>
> Also, if any of the other designers feel I have not described the intent
> correctly, please speak up!
>
> Regards,
> Eric
>
>
>
>   
>> I attached an image that shows parts of the corresponding parts of the
>> example file, the CV file, the schema file and the mapping file.
>>
>> Best,
>>   Marc
>>     
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Psidev-ms-dev mailing list
> Psi...@li...
> https://lists.sourceforge.net/lists/listinfo/psidev-ms-dev
>