Error when running "sander and PB calculation" in a charmm membrane

[David Visentin]

Hello again,
I posted a message (a few days ago) regarding PB calculations on charmm ligand membrane systems. As you pointed out, there was a hardware limitation regarding the lack of available RAM. Because of this, I switched to our local HPC cluster but now there is a different error. I'm quite new to HPC aside from running Gromacs simulations so there is a high chance I missed something.

Output error from *.log is the same as last time:
----------------------------------------------------------------------------------------------------------------------------------------------
[ERROR ] CalcError /home/dvisentin/.conda/envs/gmxMMPBSA/bin/sander failed with prmtop COM.prmtop!

If you are using sander and PB calculation, check the *.mdout files to get the sander error
.
      Check the gmx_MMPBSA.log file to report the problem.
----------------------------------------------------------------------------------------------------------------------------------------------
On the other hand, the *.mdout is different:
----------------------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------
  4. RESULTS
--------------------------------------------------------------------------------

 POST-PROCESSING OF TRAJECTORY ENERGIES
Cpptraj Generated trajectory
minimizing coord set #  1
PB bomb in pb_atmlist(): MAXNEI too short
----------------------------------------------------------------------------------------------------------------------------------------------

I tried a different combination of mmpbsa.in and slurm settings but nothing seems to work. If needed I'll upload all the files on a drive and provide a link.

Thank you in advance.

Kind regards,
David

[marioe911116]

there is a similar error reported in amber mailing list (http://archive.ambermd.org/200604/0063.html)

it seems one of your cutoff values is too high and you need to increase the maximum number of neighboring atoms per atom (MAXNEI), although I will recommend this if you are absolute sure that the cutoff you are using makes sense...

could you please, attach the mmpbsa.in file you are using?

[David Visentin]

Thank you for the response. Where would I find this option?

Of course. 

[Mario Ernesto]

well, the pb_def.h file resides in:

AmberTools/src/pbsa folder

after you change the "MAXNEI" to a larger number you will need to recompile the code... if this is not an option because you are using a public HPC, then consider sending the files you are using to see if we can provide a solution...

regards!

--
You received this message because you are subscribed to a topic in the Google Groups "gmx_MMPBSA" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/gmx_mmpbsa/UqImUx8hCI0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to gmx_mmpbsa+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/ef03e218-216f-44c9-8591-74a534c0edaen%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[David Visentin]

I'm using the gmx_mmpbsa through miniconda. Is it possible to change the setting and recompile in this situation? Again sorry for the stupid questions.

Best regards!

[Mario Ernesto]

I think in this case the ambertool compilation is better than using the miniconda, actually not sure if this can be done with the miniconda installation... have you tried with a  different "cutnb" cutoff, let's say 8?

You received this message because you are subscribed to the Google Groups "gmx_MMPBSA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gmx_mmpbsa+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/350a5c9e-ed0b-4558-a2dd-a56ecc406e42n%40googlegroups.com.

[David Visentin]

This actually did something, but another error occurred.

the *.log files look the same but *.mdout now reports this:

----------------------------------------------------------------------------------------------------------

  4.  RESULTS
--------------------------------------------------------------------------------

 POST-PROCESSING OF TRAJECTORY ENERGIES
Cpptraj Generated trajectory
minimizing coord set #     1

  NB-update: atom-based nb list  4670604


======== Setting up Grid Parameters ========
Using bounding box for grid setup
 Bounding Box at level:           1
 Bounding Box Center:      97.000    98.000    78.000
 Xmin, Xmax, Xmax-Xmin:    21.950   171.820   149.870
 Ymin, Ymax, Ymax-Ymin:    23.750   172.690   148.940
 Zmin, Zmax, Zmax-Zmin:    22.080   133.590   111.510
 Grid dimension at level     1   384  384  288
 Grid origin corrected at level     1      0.500     1.500     5.500
------- VMD goodie --------
#First Level Geometric Box

draw materials off 

draw color green
draw line "    0.500    1.500    5.500" "  192.500    1.500    5.500"
draw line "    0.500    1.500    5.500" "    0.500  193.500    5.500"
draw line "    0.500    1.500    5.500" "    0.500    1.500  149.500"
draw line "  192.500    1.500    5.500" "  192.500  193.500    5.500"
draw line "  192.500    1.500    5.500" "  192.500    1.500  149.500"
draw line "    0.500  193.500    5.500" "  192.500  193.500    5.500"
draw line "    0.500  193.500    5.500" "    0.500  193.500  149.500"
draw line "    0.500    1.500  149.500" "  192.500    1.500  149.500"
draw line "    0.500    1.500  149.500" "    0.500  193.500  149.500"
draw line "  192.500  193.500  149.500" "  192.500  193.500    5.500"
draw line "  192.500  193.500  149.500" "    0.500  193.500  149.500"
draw line "  192.500  193.500  149.500" "  192.500    1.500  149.500"
#2nd Level Geometric Box
draw color blue
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
draw line "    0.000    0.000    0.000" "    0.000    0.000    0.000"
--- End VMD goodie --------

======== Setting up Solvent Accessibility Data ========
Setting up working radii
Number of SA srf points exposed161166

SA Bomb in circle(): Stored surface points over limit******

[Mario Ernesto]

try changing "maxarcdot" parameter to 1500 and see if the error persists

To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/99ba7d71-e39c-4124-bef5-50619a8419dfn%40googlegroups.com.

[David Visentin]

Unfortunately, since I'm the only one who would use gmx_mmpbsa the administrator can't go forth with the installation.

Kind regards,

David

[David Visentin]

The same error occurred.

[Mario Ernesto]

it's a little bit odd... we have worked with big systems in the past without issues... please, consider sending us the files you re using to examine more in detail what's going wrong

regards!

To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/9078c841-c00f-435c-abe4-34ccde29e822n%40googlegroups.com.

[David Visentin]

I'm sorry when you said " try changing "maxarcdot" parameter to 1500" did you mean "15000"?

For some reason, something is happening now!

[Mario Ernesto]

I suggested 1500 because is the default option... But let's see if 15000 works well

maxarcdot (Default = 1500)

Number of dots used to store arc dots per atom.

To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/54bc7b7c-6eb4-436d-8ec4-11cee37fbf7fn%40googlegroups.com.

[David Visentin]

I tried 1500 again and still the same error. For some reason,  "maxarcdot" set to 15000 works. I'll respond again after it ends to report it run well. Should I still send the files?

[Mario Ernesto]

If you can send the files so I can see what's going on with this maxarcdot thing, that would be great

To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/c3a722e6-bc76-4953-a96e-2feb2a6fccb9n%40googlegroups.com.

[David Visentin]

I can't thank you enough for your help! 

Of course! Here is the link to the drive with all the files:

https://drive.google.com/drive/folders/1GPiKu4KCj4f0Cii6y6mYcCpLbn9fy-IG?usp=sharing

Warm regards,

David

[Mario Ernesto]

Hi there! I checked the amber manual and found some info:

maxarcdot=1500 actually means automatically determine number of arc dots required for solvent accessible surface, might grow too large to fit machines with less available memory. Please assign it to 4000~7000 and see if it fits into your computers.

I did the test and 7000 works fine (and presumably 15000 -the one you are using- works equally well), probably because of how large your system is...

hope this helps!

To view this discussion on the web visit https://groups.google.com/d/msgid/gmx_mmpbsa/791d73e5-8f7e-4843-8e4e-24d5984f7e03n%40googlegroups.com.

[David Visentin]

Hello again, the job is still running normally.

For other combinations of "maxarcdot", => 6000 seems to be the sweet spot!

Thank you, it helps a lot! 

Warm regards,

David

[marioe911116]

Great!