Problem exporting data processed in MS-DIAL to GNPS

[Anauaralima]

Hello, I am using version 5.5.24 of ms-dial to process the data and generate the mgf file and the txt file of the feature detection and GNPs has given an error when inserting the data generated by ms-dial and cannot generate a network. . There is a new way to export data directly in these new versions. In versions 4, there was the possibility to select to generate the GNPS table and now this possibility no longer exists. I don't know if this function has been replaced.
Thank you in advance!

[diosphenol]

GNPS is not suitable for MS-DIAL 5 data, and you can try GNPS2 (https://gnps2.org). However, I can find some features with RT, M/z as zero imputed into the table and the network. I believe the input of @mwang87 will be valuable. Here is example GNPS FBMN for MSDIAL 5 data https://gnps2.org/status?task=645ec447ff344b7a82f12083537167f7

[diosphenol]

For example, extra nodes with RT, m/z of zero, are attached.

[diosphenol]

P.S. These data are MSClean filtered, so the alignment IDs are not in a numerical sequence (if that may help) @mwang87. Many thanks Ming in advance.

[mwang87]

GNPS2 should handle new formats for MSDIAL5 - there were updates that GNPS1 cannot handle. I think in general with MSDIAL5 we shoudl be renumbering the scans - so sequential scan numbers in the original file are not strictly necessary.

[aliciacardoso-cmd]

I use MSDIAL5 to generate the molecular network in GNPS2. But I believe it is not compatible in the same way. There is an error when exporting the ion table from MSDIAL5. In version 4 of msdial, I can generate my workflow normally. However, the new version does not have the “GNPS table” option that it had previously. Something in the generated file, which in my case replaces the table called “MS/MS included,” leads to an error where several m/z and RT values appear as zero. I request that you look into this problem or help us with a tutorial on how to correctly export the aligned data. I loved version 5 and the possibility of using more than one database, but unfortunately I always go back to version 4 because msdial5 does not allow me to proceed to the molecular network. Please help us!

[mwang87]

Great! Can you send the job url?

…

On Fri, Aug 22, 2025 at 8:55 AM aliciacardoso-cmd ***@***.***> wrote: I use MSDIAL5 to generate the molecular network in GNPS2. But I believe it is not compatible in the same way. There is an error when exporting the ion table from MSDIAL5. In version 4 of msdial, I can generate my workflow normally. However, the new version does not have the “GNPS table” option that it had previously. Something in the generated file, which in my case replaces the table called “MS/MS included,” leads to an error where several m/z and RT values appear as zero. I request that you look into this problem or help us with a tutorial on how to correctly export the aligned data. I loved version 5 and the possibility of using more than one database, but unfortunately I always go back to version 4 because msdial5 does not allow me to proceed to the molecular network. Please help us! Screenshot.2025-08-22.125723.png (view on web) <https://github.com/user-attachments/assets/f27c10eb-8bce-495a-9dd9-cf62079403ad> Screenshot.2025-08-22.125022.png (view on web) <https://github.com/user-attachments/assets/6173e0b6-6532-46a6-88af-2d0a64576c52> — Reply to this email directly, view it on GitHub <#497 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAXSEDS7TNWCLFBPPJ4DWL3O44QRAVCNFSM6AAAAACESLC4XWVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTIMJZGEZTINI> . You are receiving this because you were mentioned.Message ID: <systemsomicslab/MsdialWorkbench/repo-discussions/497/comments/14191345@ github.com>

[aliciacardoso-cmd]

Sure! Thanks for your help!
Link: https://gnps2.org/status?task=e10ee534da384a5d8ef58e1e4aace45f#

[froz9]

I am using this app to convert mgf files from MS-DIAL to GNPS and Sirius. Try it, I found it very useful:

https://ms-dial5mgf-to-gnps-sirius.streamlit.app/

[alicialudy]

Thanks for the tip. I will certainly try it!

[mwang87]

@froz9 This is very cool. If you're interested, feel free to add to this - https://wang-bioinformatics-lab.github.io/GNPS2_Documentation/toolindex/

so others can see it and cite you.

[froz9]

Thank you. I will try to add it during the week.