Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Tom-Pope
    @Tom-Pope
    if you could let me know the common cause for this error I will trouble shoot it myself
    Mohammad Soroush Barhaghi
    @msoroush

    @Tom-Pope I looked at your system, and I believe the problem is in the parameter file.
    This problem only happens when you are trying to construct a branching point that forms a plane and this a due to mathematical limitation.
    Let's focus on the oxygen atom attached to the benzene ring in aspirin molecule. In your parameter file we have:

    Angles parameter:
    Atom Types                         Ktheta(K)         theta0(degree) 
    OR    CB    CB                      35056.0000         120.0000
    CB    CB    CB                      24227.6600         119.9770

    We have one OR-CB-CB and two CB-CB-CB angles at this branching point.
    Having 3 angles with equilibrium value near 120, it means they must be in the same plane. When you have flexible angles, if you pick an angle above 120 (let's say OR-CB-CB angle is 122), it means that sum of the both CB-CB-CB angles must be less than 238, otherwise you cannot construct such a geometry. If in your trial angle generation, you try angles with summation above 238, you will get this error message.
    To solve your problem, I would suggest to fix the angles for OR-CB-CB, CB-CB-CB, CB-CB-HCMM. Its very important to make sure when you build your molecule, the CB-CB-CB, CB-CB-HCMM angles are exactly 120. Otherwise we would get the same error. Until we find another way around to solve this problem.

    also I there is another error being generated in my Ibuprofen simulations: 'Wrapping molecules inside the simulation box 1:
    Error: Multiple atoms with zero coordinates were found'

    Please check you PDB file for simulation box 1. Sometimes packing or PSF generation failed which results in generating zero value for X, Y, and Z coordinates. If you found multiple molecules with zero coordinate, it means PSF generation failed due to possible mistake in topology file. You can look into the warning generated from vmd.

    Tom-Pope
    @Tom-Pope
    thank you very much --- got my aspirin simulations up and running and I'm sure I will find the error in the topology for ibuprofen when I get round to it
    Tom-Pope
    @Tom-Pope
    Hi I've noticed during restarting simulations that the total energy of the system is lower in magnitude post restart to the pre-restarted simulation, I've checked the contributions and the electrostatic energy reduces in magnitude by 1e+05 (for my particular case). I boiled it down to the contribution form the self interaction and correction electrostatic energies --- it seems to me that the self interactions and correction electrostatics of the framework (box0) is recomputed when a simulation is restarted with adsorbates already in the box 0 containing the framework (taken from the restart coordinate files) --- would it be possible to add in a flag to the configure file that allows you to use the pre restarted self interaction and correction to allow for simulation continuity --- this would be a very helpful feature for my style of research where adsorbate-adsorbent interactions are a main goal and simulations will have to be restarted a lot due to the size of the guests
    Mohammad Soroush Barhaghi
    @msoroush
    @Tom-Pope since there are multiple ways to define adsorbent structures, it would be very difficult to detect it and exclude it from simulation interactions. Can you use simulation box with no adsorbate, to calculate the energy associated with framework and subtract this energy from your restart or recalculated energy?
    Tom-Pope
    @Tom-Pope
    Hi, yes that's what I've been doing I just thought it might be a good idea in aid of high throughput but it's not a tough task to automate for many structures
    cheers
    Tom-Pope
    @Tom-Pope
    I have a query about dihedrals terms --- if I wan't to use a mixture of TraPPE style dihedrals, CFF dihedrals and three cosine dihedrals, is it valid to use these in GOMC converting to the charmm style format? also for the reference angles needed by charmm do I choose either 180 or 0 for dihedral forms that do not provide reference angles based on the configuration of my molecule?
    Mohammad Soroush Barhaghi
    @msoroush
    Hi @Tom-Pope. Please mention me in your text, so I get the notification. Regarding your question, TraPPE sometimes uses non standard form of dihedral as you may saw. The best and the safest method is to try to refit the TraPPE dihedral energy to Charmm style dihedral, no matter what phase shift value (0, 180) you use. This is valid as long as you reproduce the same dihedral energy as the function phi, no matter what equation form you are using.
    Tom-Pope
    @Tom-Pope
    @msoroush ah okay thank you
    Tom-Pope
    @Tom-Pope
    @msoroush Hi Mohammad, when reading CIF files into GOMC, at what decimal place are the read in charges truncated? I'm using some structures with charges up to ten decimal places, when using the HTS tool to set up my simulations it seems to be truncating the charges at six dp causing rounding errors. So I was wondering if ~GOMC also truncates the charges at six dp
    Mohammad Soroush Barhaghi
    @msoroush
    @Tom-Pope GOMC reads the charges as they printed in PSF file. So, if in PSF file it's printed with 6 dp, GOMC would read 6 dp.
    Mohammad Soroush Barhaghi
    @msoroush
    What is the net charge reported in log file?
    Mohammad Soroush Barhaghi
    @msoroush
    @Tom-Pope I had to truncate the charges to detect the unique charges. Two atoms with less than 0.000001 difference in charge is meaningless
    Tom-Pope
    @Tom-Pope
    @msoroush the net charge reported in the log file for one of my MOFs with 15 d.p. charges (that sums to zero with these) is -0.0002 --- I figured that GOMC would be reading them in to 6 dp but I'm not sure how to get around this issue without arbitrarily removing charges from atoms
    Mohammad Soroush Barhaghi
    @msoroush
    In molecular mechanics simulation the uncertainty in the results are high. So if you even read the charge with 15 dp, you don’t see changes in results. Most force fielda are using 3,4 dp for charges
    If you still want to read the charge with 15 dp, I can help you to modify the script to generate PSF file with 15 dp, even though it’s not standard format of PSF file
    Tom-Pope
    @Tom-Pope
    @msoroush okay -- thanks for the offer but I think I should be fine ---- the charges I compute for my MOFs will always be to a higher degree of accuracy than what will be read into the programme but I will run some tests anyway just to be vigilant that small net charges on my system aren't skewing my results
    Tom-Pope
    @Tom-Pope
    @msoroush Hi Mohammad, I had an old version of GOMC (where the restart procedure got stuck in a loop) compiled on one of our research computers --- I've re-pulled the development branch from github but the ./metamake.sh spits out errors and doesn't finish compilation
    Screen Shot 2019-12-04 at 13.19.53.png
    this is the error that compiling gives after scanning for the dependancies for the NPT ensemble
    I also noticed that there is now a metamakeMPI.sh file --- are you/have you made GOMC an MPI parallelised code?
    Tom-Pope
    @Tom-Pope
    also read in the GOMC paper that work will be done into implementing continuous fractional MC --- I'm curious of how far off you are on implementing this algorithm?
    Mohammad Soroush Barhaghi
    @msoroush
    @Tom-Pope sorry for the compiling error. We are in the process of implementing parallel tempering
    The MPI-OpenMP is the first step toward Parallel tempering implantation, where each replica runs at separate node. Also it allows user to run multiple NVT, NPT, and GCMC simulation with only one config file, where each simulation runs at separate node.
    Mohammad Soroush Barhaghi
    @msoroush
    One of our developers fixed the problem and once we merge it to the development branch, I will let you know. @YounesN
    @Tom-Pope Regarding implementation of CFCMC, I was working on that for a while, it’s up and running but not stable. For complex system such as heterogeneous system, the bias doesn’t work properly and system get stuck in intermediate states.
    GregorySchwing
    @GregorySchwing
    @Tom-Pope Please pull the development branch and retry compilation.
    Mohammad Soroush Barhaghi
    @msoroush
    @go2432 I merge it but I think it needs @YounesN approval for merging
    Mohammad Soroush Barhaghi
    @msoroush
    @Tom-Pope it seems that the fix has been merged. Would you please pull the development branch and compile it again. Let us know if you faced any problem
    Tom-Pope
    @Tom-Pope
    @msoroush @go2432 re-pulled the development branch from github and standard CPU executable compilation works fine, however I have not built the GPU executables, cheers
    Tom-Pope
    @Tom-Pope
    @msoroush @go2432 From my last message when I compiled GOMC I only tested to see if the executables were built, today I noticed that they are not functioning properly. I've tested the GCMC, NVT and NPT executables (not GEMC as I didn't have the relevant files set up) on a version of GOMC a pulled and compiled today. For each executable, when ran, the error 'segmentation fault: 11' is produced and only the the out.log file is produced ( I will attach). I tested the simulations with an older build of GOMC and everything ran fine, cheers
    GregorySchwing
    @GregorySchwing
    @Tom-Pope Thank you for pointing out that bug. If you repull, it is fixed.
    Micah Welsch
    @micahwelsch
    Hi there! My name is Micah and I've been using GOMC here and there for a while. I'm trying to simulate gas adsorption in an amorphous polymer system and keep getting segmentation faults after the dihedrals of the force field are read in. I've checked to make sure that all of the dihedral parameters are present. The only thing I can think of that's causing the issue is that I didn't specify the parameters correctly. I'm attempting to use the OPLS forcefield and set it up similarly to how it was previously suggested in these forums but I may be doing it incorrectly. Or the issue lies somewhere else entirely. Attached are the relevant files.
    Thanks!
    Jeffrey Potoff
    @jpotoff
    @micahwelsch thanks for sending. We'll take a look at what you have here and get back to you ASAP.
    @micahwelsch can you do us a favor and open an issue about this on Github? Even if something isn't right in your input files, GOMC should not be segfaulting on you. What version of GOMC are you running?
    Micah Welsch
    @micahwelsch
    @jpotoff Thanks for the quick response! I opened an issue on Github. I'm using version 2.51.
    Mohammad Soroush Barhaghi
    @msoroush

    @micahwelsch I looked at your system, It seems that the psf file is corrupted. I could not visualize it in vmd and I had to look up the bond list in GOMC output and create the bond inside vmd using

    set bonds {{0 4} {0 12} {0 1} {1 15} {1 2} {2 14} {2 3} {3 5} {3 4} {5 6} 
    {5 7} {7 8} {8 9} {9 10} {10 11} {11 45} {12 13} {12 33} {16 20} {16 17} 
    {16 27} {17 30} {17 18} {18 29} {18 19} {19 20} {19 21} {21 22} {21 23} {23 24} 
    {24 25} {25 26} {26 33} {27 28} {27 31} {31 32} {34 38} {34 35} {34 45} {35 48} 
    {35 36} {36 47} {36 37} {37 38} {37 39} {39 40} {39 41} {41 42} {42 43} {43 44} 
    {44 49} {45 46} {49 50} }
    
    set all {atomselect top all}
    topo clearbonds -sel $all
    topo setbondlist $bonds -sel $all

    But even after bond setting, your system looks very unusual. Anyway, after debugging, I notice that you specify bonds between {0 4} {0 12} {0 1} which creates a branch molecule. However, you only defined 2 angles instead of 3 {1 0 4} and {1 0 12}. So, we are missing the third angle {4 0 12}. This error comes from the fact that {4 0 12} angle is not provided in PSF file. Please use auto angle dihedral in your topology file, so the angle and dihedrals get generated automatically. We should also provide more meaningful error for user to fix their system.

    Micah Welsch
    @micahwelsch

    Thanks for the help earlier with my polymer issue! I ended up getting passed the previous problem but I'm running into a new one now.
    Since upgrading to the newest version of GOMC I've been getting a "Cannot Construct Angle X Y Z!" error message for a specific angle in my polymer system and I can't quite figure out a fix.

    I'm attaching the necessary files for a test calculation in the NPT ensemble. Some of the pathing may need updated, however.

    GregorySchwing
    @GregorySchwing
    Screenshot-from-2021-03-29-08-41-43.png
    GregorySchwing
    @GregorySchwing

    Hello Micah,

    I have spent some time looking into your issue. GOMC does support sampling molecules like the one you posted. That is the angle that the error seems to stem from, which is similar to the ring of a proline amino acid. I'm curious about the source of your pdb and psf files.

    Could you please open an issue on our github: https://github.com/GOMC-WSU/GOMC/issues/new/choose

    Also, please attach your ./ptf-molecule.top file. If possible, provide a SMILE string of this molecule, as we recently developed MosDef support for directly generating psf/pdb files from a smile string.

    Thank you for using GOMC!

    Sincerely,
    Greg

    Micah Welsch
    @micahwelsch

    Thanks for the reply @GregorySchwing

    I wrote the initial PDB file with Avogadro but edited the atom labels to match the identifiers in a topology file that I wrote from scratch. I filled a box with several molecules with packmol and then fed that pdb file into a simple .tcl script that I ran with VMD.

    I was able to get this system to run previously with this method but I'm open to having made an error.
    Sorry I should have included those additional files. Here's the same zip file but with the topology, single molecule pdb, and packmol input/output files I used to get the START_BOX_0.pdb/psf. I'll send all of this to a ticket on github.

    Best,
    Micah

    Micah Welsch
    @micahwelsch

    Sorry, also I here is the SMILE code for this pentamer of this polymer I'm using as my example. I believe I acquired this correctly.

    O=C(O)c5ccc(C(=O)OCCCOC(=O)c4ccc(C(=O)OCCCCOC(=O)c3ccc(C(=O)OCCCC(=O)OCc2ccc(C(=O)OCCCOC(=O)c1ccc(C(=O)OCCCO)o1)o2)o3)o4)o5