No chemical information in qcxms.out #79
Comments
|
Hello, that sounds very interesting, thanks for sharing! Normally there would be an automatic creation of a If you want more control, you could add new lines there as well, e.g. increasing the number of trajectories, electronic impact energy, etc. (see the link above on how the corresponding commands are called) @gorges97 : can you check why the input file is not created automatically? |
|
Hello, this is a strange error and unfortunately I cannot reproduce it. |
|
Hi dear @JayTheDog, Sorry for the silence and late, I just got back from vacation. I previously executed the "qcxms" command without providing any "qcxms.in" file because as described your description https://xtb-docs.readthedocs.io/en/latest/qcxms_doc/qcxms_run.html# input-keywords-in-qcxms-in assumed that EI calculation would be conducted. I just now re-executed the "qcxms" command with "ei" as a parameter in "qcxms.in" file. This takes several hours of running. thank you for help |
|
Hi dear @gorges97, Sorry for the silence and late, I just got back from vacation. I ran the "qcxms --prod" command in one of the TMPQCXMS subfolders in which the following files had been generated (qcxms.out, qcxms.in, qcxms.start, start.xyz). After running the script the following ones were generated (1.1.xyz, 1.2.xyz, 1.3.xyz, ion.out, neutral.out, xtb.last, MDtrj.98.0.1, and qcxms.res), but I notice that the qcxms.res file is empty. $ls -l Thank you for help |
|
Could you please send me the file qcxms.out? The generated files seem to be OK, but the fact that the qcxms.res file is empty is really strange. |
|
For me is same, no chemical information in qcxms.out |
|
Hi dear @JayTheDog, The running of gcxms with the command 'ei' in the file "qcxms.in" (as you recommended) has just finished. Attached are the tracks of the execution (qcxms_tracks.txt). The results are identical to those issued with no file "qcxms.in" file. I add qcxms.out, wrapped_qcxms attached as txt files. wrapped_qcxms.txt NB: I noticed that the wrapped_qcxms is created temporarily when creating qcxms.out, then is automatically deleted afterwards. Is this normal? Is there a step after the protonation step before running qcxms? I am sending you an example of our data in order to help to determine the origin of bug. You will find in zip file, the struc.xyz for example. and the deprotonated.xyz, protonated.xyz, qcxms.in, coord file (obtained from crest tool). Thank you for your help |
|
Hello dear @anani-a-missinou, It seems that you are using an old pqxms script which creates a wrapped_qcxms file with the command "qcxms -prod", but it has to be "qcxms --prod" with two dashes! Please download the latest pqcxms script from the 5.2.1 release package and use it for your calculation, this should fix your problem. It is correct that the wrapped_qcxms is deleted after running pqcxms. If you want to calculate a spectrum for a protonated structure, you need to take a structure (usually the lowest in energy) from the protonated.xyz file and copy its coordinates into the input geometry file. QCxMS does not read the protonated.xyz file from CREST. |
|
Hi dear @gorges97 I installed version 5.2.1, but it killed while running. This is track. ########## QCxMS is free software: you can redistribute it and/or modify it under QCxMS is distributed in the hope that it will be useful, Cite this work as: for the CID module: with respect to negative/multiple charges: for the GFN1-xTB implementation: for the GFN2-xTB implementation: changed by input:
----- Internal program parameters ----- QC Program : xTB M+ Ion charge(charge) : 1 ------------ EI settings ----------- qc path /usr/local/bin/xtbhome directory ~/.XTBPARAM/Generating uniform velocity distribution, T= 500.0 initializing GFN2-xTB ... step time [fs] Epot Ekin Etot error #F eTemp frag. T Thanks! |
|
I just meant that you have to use a newer pqcxms script. You can also leave everything else the same and just replace the "-prod" in line 92 with "--prod" in your current pqcxms script. Of course it's also a problem that the newest version crashes for you, I'll have a look at that too. |
|
Hi dear @gorges97 The ligne 92 of pqcxms already contain --prod. I ran pqcxms v5.2.1 on files previously generated by qcxms v5.2.0 (conda version) because when I try to run qcxms v5.2.1, it does not generate qcxms.Mspec.xtb2 and qcxms.Mspec.tbxtb files and xtb.last file generated only contain only this information ""[Info] Starting GFN2-xTB calculation in tblite""", suggesting steps for equilibration is not ended. I hope the script of qcxms is the same between the 5.2.0 and 5.2.1 versions Now pqcxms gives two files tmpqcxms.res and tmpqcxms.out, added as *.txt files. Awesome! Now I'm trying to compare of calculated spectrum to the experimental spectrum (formatted as intensity in the first column and m/z in the second column). I try the command plotms -e m321.107.csv. The format of CSV is correct? He only outputted three files result.csv, results.jdx, mass.agr. Is it possible to export in image format as PNG or TIFF, ...? Thanks for your help |
|
Great! Your m321.107.csv file is not in the right data format for plotms. It has to be like the result.csv file: |
|
Hi dear @gorges97, Great, everything is fine. Thank so much ! Is it possible to directly plot the matching score on the comparison graph or at least in the mass.agr file? I got a score of 0.400, which I think, corresponds to 400 in your article. What are your recommended thresholds for judging unreliable similarities? and tips for improving fragmentation? In addition, Is there a threshold number of atoms or m/z for which you do not recommend calculating the fragmentation? because some Brassica molecules have m/z up to 1000 Da, as flavonoids are sometimes multiacylated and multi-glycosylated. Knowing that our experimental spectra come from a Bruker Daltonics UHR-QqTOF Impact II spectrometer, or UPLC-Orbitrap-Q-exactive or Orbitrap Exploris 240 for which the fragmentation experiments were performed using ESI+ and ESI- with argon as a collision gas for collision energy from 20 eV to 40 eV. Can these bellow parameters be correct? and what are your recommendations for amelioration and what are the critical parameters? For CID (counterpart of ESI+)cid for DEA (which I think is a counterpart of ESI-?)ei Thank you in advance for all of these precisions! |
|
Adding the score directly to the mass.agr with plotms is not possible, but you could add the score manually in the mass.agr file. For example in the line "@ string def "M\S+" and add the score where the M+ is printed in the plot. Yes, this is right, the maximum score printed is 1.0 which corresponds to 1000 in the publications. Usually, a score of 600-700 can be considered to be a really good agreement but the score is very dependent on the peaks with higher m/z values and thus on the relative height of the molecular peak, which can be changed by increasing or decreasing the impact excess energy (keyword ieeatm ). However, I would look more at whether the relevant peaks are predicted and only use the score as a rough guide. Regarding the system size, I would not go way beyond 100 atoms (including hydrogen atoms), as the computations are getting very expensive for large molecules. The energy settings of your CID input look good, but I would first try to find good energy settings with fewer trajectories (e.g., "ntraj 100"), as we observed that the correct energy settings can be very system-dependent, especially in CID. DEA is not the counterpart of ESI-, but of EI- (open shell ionization). For your system size, I would recommend using xtb2 first instead of DFT because it is very expensive. But since negatively charged systems are generally more difficult for QM methods, it may be necessary to use DFT. |
Hi dear developer,
Basic information
Our laboratory works on the specified metabolites involved in the ecological interactions of crops in the Brassicaceae family (Within this group, we can find crops, e.g., broccoli, cabbage, cauliflower, kale, Brussels sprouts, turnip, or rapeseed.
We are interested in your quantum-chemical electron ionization mass spectra QCxMS program to generate CID/EI spectra of known plant small molecules for which authentic references are almost absent.
My operating system: Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz 2.21 GHz, 16.0 GB (15.8 GB usable)
What happened?
The CREST protonation tool works very well and creates protonated file. First, we wanted to run EI. The first execution of the qcxms program generated two files qcxms.gs and trjM. Afterward, the second execution of qcxms program generated xtb.last, qcxms.Mspec.xtb2, qcxms.Mspec.tbxtb files and created a TMPQCXMS folder which contains several subfolders containing qcxms.in, qcxms.start, and start.xyz.
However, when running the command "pqcxms -j 12 -t 6 &", the qcxms.out file below was created and presumably does not contain chemical information.
qcxms.out.txt
Thank you
The text was updated successfully, but these errors were encountered: