[gmx-users] MPI on single quad-core

Hello I am attempting to run REMD on a single quad-core computer with gromacs 5 (MPI version). I was  hoping to run replicas on different threads...I was doing this to step up gromac scripts (make sure they run before I place them in the que!!) that I would then submit to a large "Advance

[gmx-users] mdrun_mpi performance

HelloI finally got my mpi gromacs 5.0.4 installed and running (thank you Justin!). The problem is that the mdrun_mpi is running at 0.55ns/day on my quad-core 2.8 GHz computer while on my quad-core 2.0 GHz computer it is running 6.0ns/day. On both computers I am using all threads. The only

[gmx-users] Gromacs GMX_BUILD_FFTW warning in building Gromas 5.0.4

HelloI have been attempting to compile and build Gromacs 5.0.4for a couple of weeks. I have been following the instructions (having also readGromacs 5.0 instruction in Manual..) at https://ourphysics.org/wiki/index.php/How_to_compile_Gromacs#Gromacs_5.1.1_MPI.2C_single  Specifically, I have

[gmx-users] Errors in building gromacs 5.0.4 MPI version

I got the following errors when I tried to install gromacs MPI version 5.0.4 on a Fedora 18 I wrote on the command line the following: cmake '/home/steve/Downloads/gromacs-5.0.4'-DGMX_BUILD_FFTW=ON -DGMX_MPI=ON  SCREEN PRINT OUT: -- Found MPI_C: /usr/lib64/openmpi/lib/libmpi.so  -- Could NOT

[gmx-users] CMAKE compilation error

Thanks for your response. I have attached the files you asked forThe "Cmake_Screen.output" contains, at start, my cmake input command -Steve On 9/20/16 2:52 PM, Steve Seibold wrote: > Hello > I am attempting to compile Gromacs 5.0.4 mpi version and I g

[gmx-users] libfftw3f.so problems....

Hello I am attempting to compile Gromacs 5.0.4 mpi version and I get the statement that the GMX_BUILD_FFTW was not used. That is cmake could not get fftwf_have_simd_avx and the other simd .so files. However, it did locate libfftw3f.so, but could not find simd files "in" libfftw3f.so. The

[gmx-users] Installing/Uninstalling

HelloI have Gromacs 4.5.4 installed (without the MPI enabled for multiple machines) on a Fedora 18 OS; but would like to install Gromacs 5.0.4 with MPI enabled. I want to run REMD simulations and it is my understandingthat I need MPI enabled (even for a quad-core machine) to run separated

[gmx-users] OPLS problem

HelloI have gromacs 4.5.4 on one of my machines and have utilizedit many times successfully. However, when I switched force fields to OPLS Istarted getting an error that states “number of coordinates in coordinate filedoes not match topology file”. I do not get this error with any other

[gmx-users] Umbrella sampling version 5.0

I have used the older version of Gromacs 4.x.x to do umbrella sampling; I have just installed Gromacs 5.0 and found that the tutorials on line are now for Gromacs 5.1. Is there a huge difference in the protocol from 5.0 to 5.1? I used a youtube illustration to install Gromacs. I tried to

[gmx-users] Distance Restraints

I have a two peptide system with  my second chain B having as part of its topology file (and sequence) a Zinc ion. That is, the Zinc ion is literally part of the chain so, that when I generate my itp files using pdb2gmx I get topol_proteinA.itp, topol_proteinB.itp (i.e. no topol_Ion.itp)...I

[gmx-users] Replica Exchange: Threads vs MPI

I apologize if this is a really crazy question; I just want to be sure that I understand everything. I am assuming after reading the Replica Exchange from the on-line gromacs site that you cannot run, perform this method with multiple threads on one one system. You must have several machines

[gmx-users] trjconv problems

I have been working on post modifying my trajectories using trjconv. My protein has two zinc ions, and I have used PBC (explicit waters). After a week of attempting to get my simulations to behave due to boundary effects, I have finally got my protein to not fall apart, and even managed to get

[gmx-users] Problems with trajectories

I REALLY need some help here. I have tried for several days now and looked on line to fix the situation. My protein moves around and goes outside the simulation waterbox and my Zinc ions also frequently have bonds to protein of 40 angstroms are so. Here is one of the combinations I have tried

[gmx-users] PBC problems??

My protein breaks according to viewing the traj in VMD and graphing the RMSD of the protein C-terminus I have tried all combinations of trjconv -pbc -center -box center and nothing works..I was able to get online and find a tutorial that says trjconv -pbc mol, should stop the problem, but

[gmx-users] Odd protein behavior in MD

I ran a 8 ns simulation on a small protein in a water box using PBC. I then calculated the RMS using g_rms command. When I plotted the outcome my rms graph looked like a Histogram instead of a rms plot. I then took a region of the trajectories and view them in VMD. The protein starts out ok,

[gmx-users] Martini and Gromacs 3 and 4

Hello I was under the impression that I needed Gromacs 3.3.1 to implement the Martini FF. But as I have been looking at mdp files for Martini, I am finding statements about this mdp file to be used for Gromacs 4. I must be missunderstanding something here...Can I use Gromacs 4.5.4 to run