search

phnmnl / phenomenal-h2020

Documentation and wiki for the PhenoMeNal H2020 E-Infrastructure Project
21 stars 6 forks source link

Empty results file from the workflow: Metabolomics LCMSMS CAMERA MetFrag processing and annotation #20

Open DaqiangP opened 7 years ago

DaqiangP commented 7 years ago

I used the supplied data with this workflow and it could run without error, but in the end, I got only empty result files. Can anybody check it and fix this issue? Thank you!

c-ruttkies commented 7 years ago

Hi, is this the default example? Can you change the database from KEGG to PubChem in the msms2metfrag node. KEGG is a rather small database oftentimes causing no matching candidates given the precursor information. With PubChem there should be hits. Just let me know.

DaqiangP commented 7 years ago

Hi, thank you for your answer. Now it works, but I got only eight files, each of which is one peak compared with the database. But I have more than hundred of peaks, how can I search the rest peaks?

c-ruttkies commented 7 years ago

Hi again, well this is not easy to answer as the peak picking and grouping in the workflow is dependent on the setting of the parameters of the single tools. It might be that only 8 peaks are properly annotated with adduct information which is needed for further processing, which is actually not much I have to admit. Can you provide your mzml file, so I can have a look at that example by myself? Is this a standard mix of compounds or an untargeted experiment?

DaqiangP commented 7 years ago

Hi, it was the same with the sample data. And our data is from an untargeted experiment, from which we could identify about 40 metabolites. Please find the file through Google drive.
https://drive.google.com/file/d/0B8w7CrpVCqFeTE5HaTlyeGw4Q3c/view?usp=sharing Thank you again and have a nice weekend!

c-ruttkies commented 7 years ago

Hi, I played a little bit with your dataset and it seems the problem is related to a shift in the precursor masses triggered for MS/MS compared to the masses we find in MS1. The shift is that high that we can simply not map the MS/MS spectra to the corresponding MS1 features by mass. Do you have an example compound from that you know it's in your data and has a MS/MS spectrum? Thanks in advance. Christoph

DaqiangP commented 7 years ago

Hi Christoph, thank you for your answer. I have checked the data and found out that something was wrong with the calibration. And then I recalibrated the masses and generated a new file. It seems that the new file works in Galaxy. There were over 7000 result files generated. Unfortunately, an error occurred in the workflow and I could not see the results. It would be nice if you can look at the new file. https://drive.google.com/open?id=0ByrBWp68y7ULQmEwNENqYmViOWM Some examples that have MSMS spectra are Glutamate, Phenylalanine and many more if you need them. Thank you! Daqiang

2017-10-12 10:08 GMT+02:00 c-ruttkies notifications@github.com:

Hi, I played a little bit with your dataset and it seems the problem is related to a shift in the precursor masses triggered for MS/MS compared to the masses we find in MS1. The shift is that high that we can simply not map the MS/MS spectra to the corresponding MS1 features by mass. Do you have an example compound from that you know it's in your data and has a MS/MS spectrum? Thanks in advance. Christoph

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/phnmnl/phenomenal-h2020/issues/20#issuecomment-336053503, or mute the thread https://github.com/notifications/unsubscribe-auth/AelDZoj045BFEadb9DuVZutiShHK368mks5srcj4gaJpZM4Pa4qV .