Skip to main content
Topic: GNPS export function (Read 442 times) previous topic - next topic

GNPS export function

Dear MS_dial developer,

I have a question about the GNPS export function.

When I use the "export GNPS" function, I have:
- the peak table
- the MGF file
I do not understand why I don't find all the IDs of the peak table in the MGF file. I took an example here (Présentation1.pdf) where I have the IDs 9,10,11,15, 26,34 which are present in the peaktable. However in the MGF file, the ID 15 and 26 are not present. If it is because IDs 15 and 26 are not MSMS data, can we create an MGF file corresponding exactly to the peak table?

Best regards community,

Re: GNPS export function

Reply #1
In addition to my question from yesterday, can we export a * .MGF file and a peaktable * .csv or * .txt using spectra from the "MS2 acquired" filter?

Re: GNPS export function

Reply #2
Hi Sebas,

I made the export function for GNPS with the suggestions from GNPS developers.
Therefore, what I made is what GNPS requires on the server.
https://ccms-ucsd.github.io/GNPSDocumentation/featurebasedmolecularnetworking-with-ms-dial/

If you wanna get all of required information with a good computational skill, "mztab-M" format export should be a great option to retrieve the standardized metadata (available from version 4.10).


--can we export a * .MGF file and a peaktable * .csv or * .txt using spectra from the "MS2 acquired" filter?
Sorry, there is no function yet. It should be in the to-do list for the future development.

Thanks,

Hiroshi

Re: GNPS export function

Reply #3
Hi Hiroshi,

would it be possible to add to export table, also a column with the list of sample (classes) where the ion was found?
That is how also GNPS creates the output table. In that way it is possible to make nodes in the newtork that are colored as "pie-charts", so split and colored by the sample class.

I did not try the mz-Tab export yet, maybe you already have this feature there, but it would be good to have also in the other export tables.

Thanks!

Stefano 

Re: GNPS export function

Reply #4
Hi Stefano,

is your image something like attached file?
This figure is the similar image of Figure 2 of the following paper.
https://www.nature.com/articles/s41592-019-0358-2

Maybe, this kind of pie chart can be created by several idea/metadata. ion abundances, ion existences, etc...
EnhancedGraphics is a good add-on on the cytoscape. And actually, you can export such a metadata (for example, ion existence) by Export-> Alignment result export -> Check "Peak ID".
In the peak id matrix, the value will be negative if the peak is not detected in the sample, and the positive value of peak ID means the peak ID of each sample file.

Please let me know if it makes sense for you.

Hiroshi


Re: GNPS export function

Reply #5
Hi Hiroshi,

yes exactly what I meant. I created these kind of networks / pie-chart node visualization in Cytoscape before, but by using GNPS exported metadata.
I think one can use the metadata from the Peak-ID matrix as you suggest.  But if possible, it would be great to add a summary to this matrix, that would report for each ion/peak in which "Class" of samples it was found. Say, if you have 3 classes, a peak may be found e.g. in Classes 1,2 , another in Classes 1,3, and another in 2,3.
GNPS exported metadata gives you a similar summary list of which samples your peak was found , as  in the defined "Group" (e.g. G1, G2, G3..), see here at min 6:00 https://www.youtube.com/watch?v=ujYR0Hugb2M&t=374s

Thanks

Stefano