Creating .sh for HPC submission of ORCA jobs

[nholub]

How can I adapt the input prep procedure for ORCA submissions on an HPC using SLURM *.sh submission files? I typically use 'sbatch *.sh' to run ORCA's .inp/.cmp jobs via a scratch folder. Can QPREP create a *.sh file for each *.inp, or should I somehow modify my attached standard *.sh file to run all the *.inp files that were created in series?

tempol_sh.zip

Originally posted by @nholub in #157 (comment)

[jvalegre]

Hi,

I normally use a SH script that runs the whole workflow AQME + Gaussian/ORCA + GoodVibes in the same script. I'm attaching an example called AQME_script.txt, but it should be AQME_script.sh (GitHub didn't allow me to upload the SH script directly):

• The "### Run QM calculation." section includes the function that executes de QM software, Gaussian in this case but it should work the same way for ORCA -> PATH/orca FILE.inp > FILE.out

• The "### Wrapper for batching and running jobs" section iterates over all the QM input files. In this case, the code iterates over COM files with *.com, but you can change it for *.inp to iterative over ORCA inputs.

Hope this helps!

AQME_script.txt

[jvalegre]

@nholub please, let me know if the issue is fixed and, if so, close this issue.

[nholub]

I've made a little progress on my .sh file for submitting ORCA jobs to the HPC but still running into some issues. CSEARCH works great but for QPREP I keep getting the error of "mem_orca" not being a defined variable (slurm submission script and all output files attached, see "spr2v_ox.e2386185"). I've tried working mem_orca it into the "rdkit_orca_script.sh" script in various ways but nothing seems to work. It could be a simple coding fix... Can you also check the rest of this sh file? I'm trying to adapt the AQME_script.sh you sent for submission on my HPC but I'm not sure if it will work. I can send you an update once the mem_orca issue is resolved. Thanks.

…

On Sat, Jul 1, 2023 at 9:49 AM Juanvi Alegre-Requena < ***@***.***> wrote: @nholub <https://github.com/nholub> please, let me know if the issue is fixed and, if so, close this issue. — Reply to this email directly, view it on GitHub <#162 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BANVCHTJ2IPGY225DE7TGTLXOBBG5ANCNFSM6AAAAAAZGVYWHM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jvalegre]

hi @nholub - I can't find the attached file, try to upload it from the thread of the issue on GitHub

[nholub]

orca_sh_test.zip

[nholub]

I uploaded the file to the thread on GitHub

…

On Fri, Jul 14, 2023 at 4:26 AM Juanvi Alegre-Requena < ***@***.***> wrote: hi @nholub <https://github.com/nholub> - I can't find the attached file, try to upload it from the thread of the issue on GitHub — Reply to this email directly, view it on GitHub <#162 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BANVCHS2OR5OVJFUCQUD35DXQENEXANCNFSM6AAAAAAZGVYWHM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jvalegre]

hi @nholub - sorry for the delay, we've been finishing the ROBERT program and I couldn't check the issues from AQME.

The problem of your script is that the memory has to be specified as 20GB instead of 20000 in "maxcore=20000". I'll add a patch in the next version (1.5.2) to accept memory values with no GB or MB included.

I checked your script, and the CSEARCH and QPREP commands worked on my computer (you might need to change the ORCA qm_input to include separated lines, maybe @turkiAlturaifi can help here).

[turkiAlturaifi]

Hi @nholub Hi @jvalegre - for the QPREP qm_input command, the following works for me in order to start a new line:

python -m aqme --qprep --program "orca" --files CSEARCH/SPr2V_ox_rdkit.sdf --qm_input "B3LYP D3BJ def2-TZVPP def2/J opt
%geom
end" --mem "1" --nprocs "1" --suffix "GP"

or

python -m aqme --qprep --program "orca" --files CSEARCH/SPr2V_ox_rdkit.sdf --qm_input "B3LYP D3BJ def2-TZVPP def2/J opt"$'\n'"%geom"$'\n'"end" --mem "1" --nprocs "1" --suffix "GP"

in this option, the line is closed with the quotes then $'\n' is added, then the new quotes for the second line and so on.

tested on mac and linux