Using CASMI Data in GNPS

[Arpana Vaniya]

Hi 

I'm trying to use the CASMI 2016 & 2017 mgf files in GNPS, has anyone tried this before? For the first example, I'm not getting any hits for molecular networking or library search. 

This is the job I'm referring to https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=d138a8b5949d4f3b9384c8335fcc3a19

I have also attached the corrected mgf file as the version on CASMI website is not formatted correctly. Is the problem related to the mgf format? Challenge 1 is from CASMI 2017 is UQGVUYNHDKMLSE-UHFFFAOYSA-N (Dihydrogenistein).

Any help would be appreciated.

Thank you,

Arpana 

======== Delete this guidelines before posting your message ========

1. FIRST, Make sure your questions was not addressed previously. PLEASE SEARCH PREVIOUS POST IN THE GNPS FORUM,

2. Examine the GNPS documentation: https://bix-lab.ucsd.edu/display/Public/GNPS+Documentation+Page

2. If you still have a question/bug, provide the GNPS URL link of jobs you are discussing about.

3. Specify the mass spectrometer type and manufacturer, as well as the procedure for data conversion to mzML/mzXML.

4. If this relevant for the issue/discussion, indicate the experimental conditions (chromatography, ionisation mode, MS/MS acquisition parameters, ...).

======== Thanks ! ========

[Arpana Vaniya]

I also received this message - There was an error retrieving the result data for block "main" of workflow type "METABOLOMICS-SNETS-V2".

Any idea how to resolve this problem?

Thank you!

[Mingxun Wang]

Hi Arpana,

I took a look at your data, and it appears there is only a single MS/MS spectrum in the input. Molecular networking doesn't help too much with a single MS/MS spectrum and is meant for whole LC/MS runs (from several to thousands). It'll help organize families of molecules together - to help assign putative relationships between known and unknown molecules. With only a single MS/MS spectrum there isn't much to organize. 

I'd recommend that if you're interested in working with a single MS/MS spectrum, you try out the library search - this will match to libraries of known MS/MS spectra, and the Single Spectrum Search (link) which will help you match to all public datasets in GNPS. 

Hope this helps!

Ming

[Arpana Vaniya]

Thank you Ming! That does help. The second example I tried on the library search worked! Do you think if I made an mgf of all the challenges in one file I could use the networking function? I will try it next, but if you already know that it won't work please let me know. 

Arpana 

[Arpana Vaniya]

Also one more question: Is the spectrum with the green peaks on the head to tail match the library spectrum?

Arpana 

[Pieter Dorrestein]

Yes we do not provide single spectrum annotation option for searches  of single spectra as this is not how data is collected in metabolomics  experiments. If you did targeted fragmentation of silege molecule then upload the entire file not a single processed spectrum.  However it should work in dereplication- only if you have the raw data and it is properly formatted as single spectrum then it may work-never tried but the dereplication workflow under data analysis is the one to use to get spectral matches against the database  https://gnps.ucsd.edu/ProteoSAFe/index.jsp?params=%7B%22workflow%22:%22MOLECULAR-LIBRARYSEARCH%22,%22library_on_server%22:%22d.speclibs;%22%7D

Molecular networking only works when you have many spectra (will merge identical spectra first and then looks for similarity among the merged spectra)  and then it will do both library matching and propagation within the molecular network to unknowns.

P

On Thursday, July 19, 2018 at 4:22:07 PM UTC-7, Mingxun Wang wrote:

[Louis-Félix Nothias]

Hi,

The CASMI challenges for 2014, 2016 and 2017 are already available on MassIVE as mgf files:

https://gnps.ucsd.edu/ProteoSAFe/datasets.jsp#%7B%22query%22%3A%7B%7D%2C%22table_sort_history%22%3A%22createdMillis_dsc%22%2C%22title_input%22%3A%22CASMI%22%7D

And the challenge solutions for 2014 and 2016 are available as a third party library. 

https://gnps.ucsd.edu/ProteoSAFe/gnpslibrary.jsp?library=CASMI

Cheers,

LF

On Thursday, July 19, 2018 at 3:58:35 PM UTC-7, Arpana Vaniya wrote: