MetFrag Web

[stan2019]

Hi, Metfrag web is not functioning well. Only KEGG database is working. ChEBI and HumDB not working.

Download candidates to csv not working Opening ranked candidates to check fragments match spectrum not working.

Thank you

[schymane]

I can confirm HMDB and ChEBI aren't working (error retrieving candidates), but many other features and databases are still working for me at the moment. Can you provide more detailed feedback on which examples are failing via the "feedback" options so they can reproduce the failed candidate download etc? It may be an issue with the database storing HMDB and ChEBI.

Can you try some of the local database versions (further down in the dropdown menu under Local Databases) in the interim? Does the download and fragment display work for you with these?

I'm not observing any unusual issues for calculations I'm running, I see fragments etc for results that are processing properly.

@sneumann the HMDB / ChEBI errors are also occurring on beta. :-(

[schymane]

Update: some things need rebuilding and @sneumann hopes things will be better on Monday to fix the ChEBI / HMDB issues.

[stan2019]

Dear MetFrag Team,

Sorry, I could not send the message using the Feedback option. Feedback could not be sent.

Thank you very much for providing the Metfrag web tool for us to use. It is a very useful tool, without it, I would not be able to process my data. I have been using it since 2020 and only in the last few weeks I noticed some issues. I don't know what happened, I checked on Chrome and Brave browser.

Under database settings

1. HMDB and CheBI give error fetching candidates
2. KEGG, PubChem and local database HMDB4_23Aug19 working well

Under candidate Filter & Scoring Settings

1. Please check if Spectral Similarity is (MoNA) is working. I manually search MoNA website with my data and pick up candidate, but Fragmentation results did not give any score for MoNA. It use to work well previously.

Under Results

1. Download Results to CSV – ERROR when downloading candidates
2. If I process candidates from KEGG database and I click the KEGG identifier code it does not open in KEGG website. I get error message WebDBGET Entry Retrieval Failed. It work perfectly before.
3. If I processed candidates from local local database HMDB4_23Aug19, no link is activated for HMDB4 for me to click on the HMDB identifier code and look up the candidate. Previously, if I process candidates from HMDB database (not local database) the link activates and I can click on the identifier code and it will open in HMDB website with further details on the candidate.
4. In the results table, under Details – Fragment open up in a new window with no spectrum plot. The download to CSV is also not working.

Please you can use just the initial data in webtool to check.

Thank you.

Stanley

Sent from Mailhttps://go.microsoft.com/fwlink/?LinkId=550986 for Windows

From: Emma @.> Sent: February 23, 2022 9:03 AM To: @.> Cc: @.>; @.> Subject: Re: [ipb-halle/MetFrag] MetFrag Web (Issue #22)

Update: some things need rebuilding and @sneumannhttps://github.com/sneumann hopes things will be better on Monday to fix the ChEBI / HMDB issues.

— Reply to this email directly, view it on GitHubhttps://github.com/ipb-halle/MetFrag/issues/22#issuecomment-1048876579, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AL5YQ3NNE7J754TGZFKRGEDU4TZKNANCNFSM5PBQUVJA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub. You are receiving this because you authored the thread.Message ID: @.***>

[schymane]

Thanks for the extra details!

Regarding MoNA, which setting are you using? Exact Spectral Similarity, or Spectral Similarity?

I use the former all the time, the latter rarely. The MoNA version is slightly older than the MoNA website (we are attempting to update).

Good to know you were testing just with the regular data, then we can attempt to reproduce. As I just posted, there are some database issues that will require a rebuild (pls check back Monday).

Re no hyperlink for HMDB via the Local Database option, this is a function of how the web interface autorecognises hyperlinks and identifiers, it is not easy for us to offer this via the Local Database format (it is an enhancement for the Server Databases) as this all works via automatic column header recognition in the Local Database section (and they are all "Identifier" - so attempted autorecognition would generate many broken hyperlinks).

My calcs were all performed with Firefox (also with the test data, but a different query formula) ... we will have to check if the fragment issues you see are browser related.

Maybe KEGG have changed their links ... @sneumann maybe something to check in the code? I don't use KEGG that often ... so have not noticed that ...

[stan2019]

Hello ,

For MoNA, I normally select both and for some candidates I get scoring to both in the results. I will check with FireFox too.

Thank you.

From: Emma @.> Sent: February 23, 2022 9:12 AM To: @.> Cc: @.>; @.> Subject: Re: [ipb-halle/MetFrag] MetFrag Web (Issue #22)

Thanks for the extra details!

Regarding MoNA, which setting are you using? Exact Spectral Similarity, or Spectral Similarity?

I use the former all the time, the latter rarely. The MoNA version is slightly older than the MoNA website (we are attempting to update). [image]https://user-images.githubusercontent.com/4070141/155346676-5ceb17e0-495f-433f-b04c-be002460114b.png

Good to know you were testing just with the regular data, then we can attempt to reproduce. As I just posted, there are some database issues that will require a rebuild (pls check back Monday).

Re no hyperlink for HMDB via the Local Database option, this is a function of how the web interface autorecognises hyperlinks and identifiers, it is not easy for us to offer this via the Local Database format (it is an enhancement for the Server Databases) as this all works via automatic column header recognition in the Local Database section (and they are all "Identifier" - so attempted autorecognition would generate many broken hyperlinks).

My calcs were all performed with Firefox (also with the test data, but a different query formula) ... we will have to check if the fragment issues you see are browser related.

Maybe KEGG have changed their links ... @sneumannhttps://github.com/sneumann maybe something to check in the code? I don't use KEGG that often ... so have not noticed that ...

— Reply to this email directly, view it on GitHubhttps://github.com/ipb-halle/MetFrag/issues/22#issuecomment-1048886889, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AL5YQ3KJWHHJRHFZEADOIXLU4T2PDANCNFSM5PBQUVJA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub. You are receiving this because you authored the thread.Message ID: @.***>

[sneumann]

He, we've had some server trouble over the last days, and as Emma mentioned are rebuilding. I'd hope some issues you experienced are collateral issues. Will keep you posted. Yours, Steffen

[schymane]

I can also confirm that MoNA is not currently returning any values for queries where we would expect a non-zero result for me too ;-(

[schymane]

@sneumann Re KEGG ... if I click on the identifiers it directs me to https://www.kegg.jp/dbget-bin/www_bget?cpd:C00716= which gets me to WebDBGET Entry Retrieval Failed like @stan2019 pointed out.

If I remove the equals, it works https://www.kegg.jp/dbget-bin/www_bget?cpd:C00716

[sneumann]

Hi, thanks for that excellent feedback and reports. We were able to confirm, and they seem to be collateral damage that happened when we upgraded some library dependencies. When downgrading to an older versions things work fine again (so nothing broken at KEGG or HMDB), but we can't put the old version online because said old libraries have security issues that made most mainstream news in december :-( We need to work these out and will keep this issue until those are fixed. Yours, Steffen

[schymane]

For the record (I think we forgot to tag this anywhere) - the feedback option seems to be broken in both old and new MetFrag versions at the moment ;-( as we also couldn't get it to work in either version.

[sneumann]

Hi, @meier-rene and myself fixed some (but not all) things. We now have https://msbi.ipb-halle.de/MetFragBeta/ The Fragments are back, the candidate should be quicker again (using a local PostGres database). Feedback works again. MoNA is untested. KEGG links still broken. Yours, Steffen

[schymane]

I just tried ... MoNA does not appear to work yet - still zero values for the default data and also another test case, which I just tried to submit via feedback.

The feedback appeared to send but I couldn't see any message to indicate an error or success of the feedback submission.

Can confirm that KEGG identifiers are still broken.. (hmm I thought I saw a message from Rene saying this was fixed)

[sneumann]

Hi, I can confirm the Feedback did arrive, Rene did fix the missing fragments and announced that in the wrong issue ;-) Culprit for broken KEGG link is identified, and a fix proposed. I might try that in the coming days. Yours, Steffen

[meier-rene]

Broken KEGG link is fixed in https://msbi.ipb-halle.de/MetFragBeta/. There are still issues with MoNA.

[meier-rene]

Ok, I think we are done. https://msbi.ipb-halle.de/MetFragBeta/ is running the most recent code. Its version v2.4.8 in the lower left corner.

List of issues we fixed:

• Offline MoNA support was broken
• Fragment viewer had no content
• KEGG outlinks were broken
• database upload dialog was broken
• Command line tool had logging issues
• Command line tool MoNA support was broken
• Candidate retrieval of several databases was broken

Thanks to all reporters. If I missed a problem please open a new issue. I think this can be closed.