MZmine 3.0.11-beta patch released

73 views
Skip to first unread message

Robin Schmid

unread,
Mar 9, 2022, 10:29:46 AM3/9/22
to mzmine-devel
To the MZmine community,

We just released the first patch version containing some fixes and small updates. 

Hope to see many of you at the first MZmine cafe in half an hour (5 pm CET / 8 am PST)
https://ucsd.zoom.us/j/96822012735 

Download the latest stable release:
https://github.com/mzmine/mzmine3/releases/tag/v3.0.11-beta

or the latest development build:
https://github.com/mzmine/mzmine3/releases/

Changelog:
https://mzmine.github.io/changelog.html

Retweet our message:
https://twitter.com/mzmine_project/status/1498974133371215878?s=20&t=y-EDbshvjR8IyMI-42QAsQ

For users:
Double click on a raw data file to explore data. For a quick start into MZmine, the new Batch wizard helps set up a standard LC-MS/MS processing workflow.

For macOS:
Currently, MZmine 3 lacks a signature for macOS. However, while we are working on this, users can allow MZmine in the macOS Gatekeeper protection by running the following command in the terminal from the folder containing the .app.

sudo xattr -cr MZmine.app
For developers:
In case you want to include your development in MZmine 3 or if your tool depends on MZmine outputs, check out our codebase on GitHub:
https://github.com/mzmine/mzmine3

Community contact:
Join our slack community:
https://join.slack.com/t/mzmine/shared_invite/zt-151egtpww-gXA9DR3OvaAIrewHYawhUQ

And join our weekly MZmine web café to discuss issues and ideas with the developers or other community members. The next meeting is scheduled for next week:
Wednesday, 9th March, 5 pm CET/8 am PST
https://ucsd.zoom.us/j/96822012735 
and will be joined by Tomas Pluskal, Steffen Heuckeroth, Ansgar Korf, Daniel Petras, Robin Schmid, and more.


Best regards,

The MZmine core team

Geert Goeminne

unread,
Mar 10, 2022, 6:35:15 AM3/10/22
to mzmine-devel
Hi Robin,

Thanks for the update, indeed now it's possible to open RAW files by dragging them in MZMine, great!

But, i have a strange view now, when openening the chromatogram, all traces are shown in the same plot, even when using filters for MS1 level:



I assume that MZmine does not recognize the full MS trace, which would be FUNC001.DAT, and FUNC002.DAT is de MSe trace, and FUNC003 would be the lock mass correction trace. But it looks like even more traces are in the chromatogram view, any idea how we can filter them out in MZmine to only see FUNC001 and 002 for example? I tried with the filters but that does not work.

If you want i can provide to you the RAW file if you want to have a deeper look at it.

Cheers, Geert

--
You received this message because you are subscribed to the Google Groups "mzmine-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mzmine-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mzmine-devel/b0d0176b-64a0-4432-a9af-7df4f45485cfn%40googlegroups.com.

Robin Schmid

unread,
Mar 10, 2022, 6:43:57 AM3/10/22
to mzmine...@googlegroups.com
It seems like mzmine is not recognizing the MS level in the raw file. I am not sure when someone of us will be able to take a look at this. But it would help us if you could create an issue on the mzmine3 GitHub and add an example raw file.

Thanks for testing.

Best regards,
Robin

You received this message because you are subscribed to a topic in the Google Groups "mzmine-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mzmine-devel/TplB5En8RQY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mzmine-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mzmine-devel/501933714.13130814.1646912106762.JavaMail.zimbra%40psb.ugent.be.

Tomas Pluskal

unread,
Mar 10, 2022, 7:34:20 AM3/10/22
to mzmine-devel
I think it's because of the way Waters data is structured into different "functions". It has always been a pain to deal with their data.... :-p

Tomas



steffen.h...@googlemail.com

unread,
Mar 10, 2022, 8:07:00 AM3/10/22
to mzmine...@googlegroups.com

Hello Geert,

 

Depending on how the waters import works, you could try filtering for a scan number range. Since both functions have normal chromatograms, I suspect that function 1 is loaded first and then function 2 is appended to all scans from function 1 (because the line goes back to 0 in the tic chart). So scan numbers from function 2 should start after function 1. However, I think the import via mzml is more intuitive.

 

In the table of the new raw data overview you should be able to find the scan number range that makes up the ms1 level.

 

Best regards


Steffen

 

 

Von: mzmine...@googlegroups.com <mzmine...@googlegroups.com> Im Auftrag von Tomas Pluskal
Gesendet: Donnerstag, 10. März 2022 13:34
An: mzmine-devel <mzmine...@googlegroups.com>
Betreff: Re: [Newsletter] [mzmine-devel] MZmine 3.0.11-beta patch released

 

I think it's because of the way Waters data is structured into different "functions". It has always been a pain to deal with their data.... :-p

 

Tomas

 

 

 

On Thu, Mar 10, 2022 at 12:43 PM Robin Schmid <rschm...@gmail.com> wrote:

It seems like mzmine is not recognizing the MS level in the raw file. I am not sure when someone of us will be able to take a look at this. But it would help us if you could create an issue on the mzmine3 GitHub and add an example raw file.

 

Thanks for testing.

 

Best regards,

Robin

 

'Geert Goeminne' via mzmine-devel <mzmine...@googlegroups.com> schrieb am Do., 10. März 2022, 12:35:

Hi Robin,

 

Thanks for the update, indeed now it's possible to open RAW files by dragging them in MZMine, great!

 

But, i have a strange view now, when openening the chromatogram, all traces are shown in the same plot, even when using filters for MS1 level:

 

 

 

I assume that MZmine does not recognize the full MS trace, which would be FUNC001.DAT, and FUNC002.DAT is de MSe trace, and FUNC003 would be the lock mass correction trace. But it looks like even more traces are in the chromatogram view, any idea how we can filter them out in MZmine to only see FUNC001 and 002 for example? I tried with the filters but that does not work.

Geert Goeminne

unread,
Mar 10, 2022, 9:52:39 AM3/10/22
to mzmine-devel
Hi Steffen,

Well scan range will be different for each sample I assume, so that would be difficult for several datafiles.

When removing the other functions in the raw data itselft and only leaving the information for function 1 it works:



But f course i need to duplicate the files before doing that to make sure no data is lost.

I will drop a file anyway on the github, because i think the Masslynx RAW files are quite nice from Waters, wait until you have seen the UEP files from Waters UNIFI system :)

MzML files are an option as well of course, I'm just afraid that people that have not much experinece with how such a file is built will have problems when they try to import Waters files directly.

Cheers, Geert



Reply all
Reply to author
Forward
0 new messages