Re: [gmx-users] Diffusion/PBC
Your best bet is probably to center everything on the receptor. That will prevent jumping of the receptor only, which is hopefully all you need. -Trayder On Tue, Nov 5, 2013 at 7:14 PM, Tsjerk Wassenaar tsje...@gmail.com wrote: Hi Debashis, Makes sure that the anion and receptor are together in the reference structure you use for trjconv -pbc nojump Cheers, Tsjerk On Tue, Nov 5, 2013 at 8:12 AM, Debashis Sahu debashis.sah...@gmail.com wrote: Dear All, I have an problem related to jumping trajectory. In my MD run, there is a receptor molecule which is binding with an halogen anion in water solvent. In the original trajectory, the binding between them looks fine but jumping present. To remove the jumping of the system from trajectory, I have used 'nojump' as discussed in the forum. Now I got a jump-free trajectory, but due to the diffusion here, I have observed that the anion and the receptor are far away from each other. I could not fix the problem. can any one suggest me? Thanks in advance. with regards, *Debashis Sahu* *Central Salt and Marine Chemical Research Institute* *Bhavnagar, Gujarat* *India, 364002.* -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- Tsjerk A. Wassenaar, Ph.D. -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] There is no domain decomposition for 16 nodes that is compatible with the given box and a minimum cell size of 0.826223 nm
Assuming you're using LINCS, from the manual: With domain decomposition, the cell size is limited by the distance spanned by *lincs-order*+1 constraints. Assuming a default lincs-order (4), 0.82nm seems a fairly sane distance for 5 bonds. Which means that you're probably using too many nodes for the size of your system. Hope that helps. If it doesn't you'll need to provide some information about your system. -Trayder On Thu, Oct 17, 2013 at 1:27 PM, Nilesh Dhumal ndhu...@andrew.cmu.eduwrote: Hello, I am getting the following error for simulation. I am using Gromacs VERSION 4.5.5 and running on 24 processors. Should I reduce the number of processor or the problem is in bonded parameters. If I use -nt 1 option. I could run the simulation. Fatal error: There is no domain decomposition for 16 nodes that is compatible with the given box and a minimum cell size of 0.826223 nm Change the number of nodes or mdrun option -rdd or -dds Look in the log file for details on the domain decomposition Nilesh -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Calculating the distance between protein and ligand during ligand diffusion out of the box
Both cases are 'real' ligand binding. If a drug binds, it binds. It doesn't matter how far away it comes from. Each periodic image is identical, it's the same ligand capable of making the same interactions in the same protein but approaching from a different angle. -Trayder On Tue, Oct 8, 2013 at 4:07 PM, bipin singh bipinel...@gmail.com wrote: Thanks for the reply Dr. Trayder and Dr. Justin. I have performed unrestrained MD with a ligand bound protein having surface exposed binding pocket (link of the image attached for clarification). I have used a cubic box with vectors 6.432nm and the system size was 4.117 3.878 and 4.059 (in nm). http://researchweb.iiit.ac.in/~bipin.singh/snapshot.png My doubt is how to discriminate between a real ligand binding/unbinding process and the rebinding observed due to PBC effects (i.e. when ligand diffuses out the box and a subsequent another ligand comes and bind from the adjacent periodic image, which may seen as a rebinding event during distance calculation). On Tue, Oct 8, 2013 at 6:37 AM, Trayder Thomas trayder.tho...@monash.edu wrote: With a ligand diffusing as freely as I'm assuming (you've omitted a lot of info, box size etc.) you aren't going to get PBC to play nice, although -nojump should have atleast given you a different wrong answer. Centering the system on the same point you are using to define the binding pocket (may require custom index groups) should get you something more reasonable looking. Also, it depends on the size of your protein and what you're doing but you should consider if it's even relevant whether the ligand is 2nm away or 5? -Trayder On Tue, Oct 8, 2013 at 6:53 AM, Justin Lemkul jalem...@vt.edu wrote: On 10/7/13 1:39 PM, bipin singh wrote: Thanks for the reply Dr. Justin. I have also thinking of the same possibility but to further confirm, I am sending the link for the plot of the distance between the COM of ligand binding pocket and COM of ligand molecule, please find some time to have a look at the plot and let me know if it seems a feasible behaviour during a simulation. http://researchweb.iiit.ac.in/**~bipin.singh/plot.png http://researchweb.iiit.ac.in/~bipin.singh/plot.png Looks like nothing more than random motion to me. Since you haven't told us what you're doing (unrestrained MD? pulling?), it's hard to comment further. -Justin On Mon, Oct 7, 2013 at 8:22 PM, Justin Lemkul jalem...@vt.edu wrote: On 10/7/13 10:46 AM, bipin singh wrote: Hello All, I have calculated the distance between the binding pocket of protein and the ligand molecule but due to ligand diffusion out of box, I am getting wrong distance as first it increases till 5nm and then decrease again to around 1nm during the simulation (which is not possible). I have fitted my trajectory with using trjconv -pbc mol -ur compact -center (protein) option. I have also tried the -nojump option but getting the same results for distances. Please suggest how to get the real distance without the PBC effect. It sounds like that very well could be the real distance. If the ligand diffused out, it simply becomes part of the solvent around the protein and can diffuse around freely. -Justin -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalemkul@outerbanks.umaryland.edu jalemkul@outerbanks.** umaryland.edu jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users http://lists.gromacs.org/**mailman/listinfo/gmx-users htt**p://lists.gromacs.org/mailman/**listinfo/gmx-users http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/** Support/Mailing_Lists/Searchh**ttp://www.gromacs.org/Support/** Mailing_Lists/Search http://www.gromacs.org/Support/Mailing_Lists/Searchbefore posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists http://www.gromacs.org/**Support/Mailing_Lists http://**www.gromacs.org/Support/**Mailing_Lists http://www.gromacs.org/Support/Mailing_Lists -- ==** Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II
Re: [gmx-users] Calculating the distance between protein and ligand during ligand diffusion out of the box
With a ligand diffusing as freely as I'm assuming (you've omitted a lot of info, box size etc.) you aren't going to get PBC to play nice, although -nojump should have atleast given you a different wrong answer. Centering the system on the same point you are using to define the binding pocket (may require custom index groups) should get you something more reasonable looking. Also, it depends on the size of your protein and what you're doing but you should consider if it's even relevant whether the ligand is 2nm away or 5? -Trayder On Tue, Oct 8, 2013 at 6:53 AM, Justin Lemkul jalem...@vt.edu wrote: On 10/7/13 1:39 PM, bipin singh wrote: Thanks for the reply Dr. Justin. I have also thinking of the same possibility but to further confirm, I am sending the link for the plot of the distance between the COM of ligand binding pocket and COM of ligand molecule, please find some time to have a look at the plot and let me know if it seems a feasible behaviour during a simulation. http://researchweb.iiit.ac.in/**~bipin.singh/plot.pnghttp://researchweb.iiit.ac.in/~bipin.singh/plot.png Looks like nothing more than random motion to me. Since you haven't told us what you're doing (unrestrained MD? pulling?), it's hard to comment further. -Justin On Mon, Oct 7, 2013 at 8:22 PM, Justin Lemkul jalem...@vt.edu wrote: On 10/7/13 10:46 AM, bipin singh wrote: Hello All, I have calculated the distance between the binding pocket of protein and the ligand molecule but due to ligand diffusion out of box, I am getting wrong distance as first it increases till 5nm and then decrease again to around 1nm during the simulation (which is not possible). I have fitted my trajectory with using trjconv -pbc mol -ur compact -center (protein) option. I have also tried the -nojump option but getting the same results for distances. Please suggest how to get the real distance without the PBC effect. It sounds like that very well could be the real distance. If the ligand diffused out, it simply becomes part of the solvent around the protein and can diffuse around freely. -Justin -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalemkul@outerbanks.umaryland.edu jalemkul@outerbanks.** umaryland.edu jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-usershttp://lists.gromacs.org/**mailman/listinfo/gmx-users htt**p://lists.gromacs.org/mailman/**listinfo/gmx-usershttp://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/** Support/Mailing_Lists/Searchh**ttp://www.gromacs.org/Support/** Mailing_Lists/Searchhttp://www.gromacs.org/Support/Mailing_Lists/Searchbefore posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Listshttp://www.gromacs.org/**Support/Mailing_Lists http://**www.gromacs.org/Support/**Mailing_Listshttp://www.gromacs.org/Support/Mailing_Lists -- ==** Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalemkul@outerbanks.umaryland.**edu jalem...@outerbanks.umaryland.edu | (410) 706-7441 ==** -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/**mailman/listinfo/gmx-usershttp://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/** Support/Mailing_Lists/Searchhttp://www.gromacs.org/Support/Mailing_Lists/Searchbefore posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/**Support/Mailing_Listshttp://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Distance restraints exploding system
To wrap this up for anyone who stumbles across this in the future: The 'solution' I ended up going with was to avoid using hydrogen atoms in the restraints, instead restraining to the e.g. methyl carbon and adding a fudge factor to the restraint length to account for the C-H distance. This results in slightly 'looser' restraints but actually runs stably. -Trayder On Tue, Sep 3, 2013 at 5:32 AM, Rafael I. Silverman y de la Vega rsilv...@ucsc.edu wrote: Have you tried with even less restraints? I found systems are not always stable with more than the bare minimum of restraints On Sun, Sep 1, 2013 at 7:54 PM, Trayder Thomas trayder.tho...@monash.edu wrote: It still explodes on a 0.1fs timestep, turning off P-R doesn't seem to have an impact. I've tried being gentle, slowly turning up the force constant and running for 1 ns for each value but as soon as the force constant approaches 100 it crashes. The starting structure was generated with the same restraints, so it is very close. I have tried using slightly different starting structures as well. I've tried running it with only 3 restraints (1 methyl group to 1 hydrogen with the restraints extended by 0.2nm so that all hydrogens are within the restraint distance) and I'm getting a segmentation fault (no LINCS warnings) at 100 steps. It runs fine with any combination of 2 hydrogens restrained, but as soon as I restrain the 3rd one I get a segmentation fault (this occurs with either setting for disre-weighting). So it seems to fair better with more restraints? The more I try to solve this problem the less it makes sense! -Trayder On Fri, Aug 30, 2013 at 6:02 PM, Mark Abraham mark.j.abra...@gmail.com wrote: On Fri, Aug 30, 2013 at 8:56 AM, Trayder trayder.tho...@monash.edu wrote: Hello, I am attempting to simulate a protein-ligand complex using distance restraints to match it to NMR data. The system runs stably without restraints. With restraints it tends to spit out LINCS angle warnings and blow up under most conditions. I'm attempting to use: ; Restraints disre = simple disre-weighting = conservative disre-fc= 1000 It blows up within 100 steps unless: I run on a single core (+gpu) or disre-fc = 100 or disre-weighting = equal If disre-weighting = conservative is causing extreme forces, then I figure it should do the same on 1 core. Not really. MD is chaotic. Small changes in initial conditions lead to different results. If domain decomposition is the problem, then I would think disre-weighting = equal shouldn't work either. I'm stumped... anyone got any ideas? http://www.gromacs.org/Documentation/Terminology/Blowing_Up has the usual suggestions - don't use P-R yet, try a smaller time step, make sure your system is close to the restrained regime (or be extra gentle until it is). Mark Thanks in advance, -Trayder Distance restraints excerpt: ; aiaj typeindex type’ low up1 up2 fac ; 2 symmetric hydrogens 1306 1389 1 10 1 0.0 0.548 1.0 1.0 1306 1396 1 10 1 0.0 0.548 1.0 1.0 ; Diastereotopic methyl groups 1306 1374 1 11 1 0.0 0.654 1.0 1.0 1306 1375 1 11 1 0.0 0.654 1.0 1.0 1306 1376 1 11 1 0.0 0.654 1.0 1.0 1306 1385 1 11 1 0.0 0.654 1.0 1.0 1306 1386 1 11 1 0.0 0.654 1.0 1.0 1306 1387 1 11 1 0.0 0.654 1.0 1.0 Full mdp: ; Run Control integrator = md ; simulation algorithm tinit= 0 dt = 0.002 nsteps = 50 ; ; Output Control nstxout = 20; write coordinates to .trr nstvout = 20; write velocities to .trr nstlog = 1000 ; write energies to .log nstenergy = 4000 ; write energies to .edr nstxtcout = 1000 ; write coordinates to .xtc ; ; Neighbour Searching nstlist = 10 ; update neighbour list ns_type = grid ; neighbour list method pbc = xyz ; periodic boundary conditions rlist = 0.9 ; cut-off for short-range neighbour (nm) cutoff-scheme = verlet ; ; Electrostatics and VdW coulombtype = PME ; type of coulomb interaction
Re: [gmx-users] Distance restraints exploding system
It still explodes on a 0.1fs timestep, turning off P-R doesn't seem to have an impact. I've tried being gentle, slowly turning up the force constant and running for 1 ns for each value but as soon as the force constant approaches 100 it crashes. The starting structure was generated with the same restraints, so it is very close. I have tried using slightly different starting structures as well. I've tried running it with only 3 restraints (1 methyl group to 1 hydrogen with the restraints extended by 0.2nm so that all hydrogens are within the restraint distance) and I'm getting a segmentation fault (no LINCS warnings) at 100 steps. It runs fine with any combination of 2 hydrogens restrained, but as soon as I restrain the 3rd one I get a segmentation fault (this occurs with either setting for disre-weighting). So it seems to fair better with more restraints? The more I try to solve this problem the less it makes sense! -Trayder On Fri, Aug 30, 2013 at 6:02 PM, Mark Abraham mark.j.abra...@gmail.comwrote: On Fri, Aug 30, 2013 at 8:56 AM, Trayder trayder.tho...@monash.edu wrote: Hello, I am attempting to simulate a protein-ligand complex using distance restraints to match it to NMR data. The system runs stably without restraints. With restraints it tends to spit out LINCS angle warnings and blow up under most conditions. I'm attempting to use: ; Restraints disre = simple disre-weighting = conservative disre-fc= 1000 It blows up within 100 steps unless: I run on a single core (+gpu) or disre-fc = 100 or disre-weighting = equal If disre-weighting = conservative is causing extreme forces, then I figure it should do the same on 1 core. Not really. MD is chaotic. Small changes in initial conditions lead to different results. If domain decomposition is the problem, then I would think disre-weighting = equal shouldn't work either. I'm stumped... anyone got any ideas? http://www.gromacs.org/Documentation/Terminology/Blowing_Up has the usual suggestions - don't use P-R yet, try a smaller time step, make sure your system is close to the restrained regime (or be extra gentle until it is). Mark Thanks in advance, -Trayder Distance restraints excerpt: ; aiaj typeindex type’ low up1 up2 fac ; 2 symmetric hydrogens 1306 1389 1 10 1 0.0 0.548 1.0 1.0 1306 1396 1 10 1 0.0 0.548 1.0 1.0 ; Diastereotopic methyl groups 1306 1374 1 11 1 0.0 0.654 1.0 1.0 1306 1375 1 11 1 0.0 0.654 1.0 1.0 1306 1376 1 11 1 0.0 0.654 1.0 1.0 1306 1385 1 11 1 0.0 0.654 1.0 1.0 1306 1386 1 11 1 0.0 0.654 1.0 1.0 1306 1387 1 11 1 0.0 0.654 1.0 1.0 Full mdp: ; Run Control integrator = md ; simulation algorithm tinit= 0 dt = 0.002 nsteps = 50 ; ; Output Control nstxout = 20; write coordinates to .trr nstvout = 20; write velocities to .trr nstlog = 1000 ; write energies to .log nstenergy = 4000 ; write energies to .edr nstxtcout = 1000 ; write coordinates to .xtc ; ; Neighbour Searching nstlist = 10 ; update neighbour list ns_type = grid ; neighbour list method pbc = xyz ; periodic boundary conditions rlist = 0.9 ; cut-off for short-range neighbour (nm) cutoff-scheme = verlet ; ; Electrostatics and VdW coulombtype = PME ; type of coulomb interaction rcoulomb= 0.9 ; cut-off distance for coulomb epsilon_r = 1; dielectric constant rvdw= 0.9 ; cut-off for vdw fourierspacing = 0.12 ; maximum grid spacing for FFT pme_order = 4; interpolation order for PME ewald_rtol = 1e-5 ; relative strength of Ewald-shifted DispCorr= EnerPres ; long range dispersion corrections ; ; Restraints disre = simple disre-weighting = conservative disre-fc= 1000 ; ; Temperature Coupling Tcoupl = v-rescale; type of temperature coupling tc-grps = Protein non-Protein ; coupled groups tau_t = .1 .1
Re: [gmx-users] How to apply trjconv -nojump to a part of a system
VMD might do what you want with the PBC tools plugin (installed by default). http://www.ks.uiuc.edu/Research/vmd/plugins/pbctools/ unwrap being the equivalent of -nojump Otherwise, couldn't you just view your 2 trajectories simultaneously, one with protein the other not? -Trayder On Wed, Jul 10, 2013 at 9:13 AM, Bin Liu fdusuperstr...@gmail.com wrote: Hi All, For the convenience of visualization, I need to remove the jump of one component (say a protein) of the system at the boundary. I don't need to, or say I need not to remove the jump of the other components (say a lipid bilayer), since otherwise the system will look falling apart. I noticed I can cluster a part of a system, then output all the atoms in the system in which only the part is clustered, and the other components unchanged. Does GROMACS have similar function when *-nojump* is used? If this can not be accomplished directly, is there a way to circumvent it? I figured out a way, but haven't implemented it. I can plug the coordinates of the protein treated with *-nojump* into the trajectory of the whole system which is not treated with *-nojump*. It is kind of substituting the coordinates of one component in one trajectory with the coordinates of the same component in another trajectory. Is there anyone aware of a tool or a script to do this job? Thank you very much. Regards Bin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Effect of pressure coupling frequency on gpu simulations
Thanks Mark, That really helped to clarify how everything is interacting around the verlet scheme. What statistics do you recommend examining between nstpcouple settings? Pressure/box size variation is the obvious one but I was curious whether you had something else in mind. -Trayder On Thu, May 23, 2013 at 4:18 AM, Mark Abraham mark.j.abra...@gmail.comwrote: On Wed, May 22, 2013 at 6:32 AM, Trayder trayder.tho...@monash.edu wrote: Hi all, I've been running 5fs timestep simulations successfully without gpus (united-atom, HEAVYH). When continuing the same simulations on a gpu cluster utilising the verlet cutoff-scheme they crash within 20 steps. Reducing the timestep to 2fs runs smoothly, however I noticed the message: Making this change manually led to crashing simulations as nstcalclr, nsttcouple and nstpcouple default to the value of nstlist. After defining them all separately I was able to determine that the simulation exploding was dependent entirely on nstpcouple and by lowering it to 5 (from the default 10) I was able to run simulations at a 5fs timestep. So, my questions: Is lowering nstpcouple a legitimate solution or just a bandaid? P-R does not cope well with situations where the box size changes enough (e.g. you should normally avoid it during equilibration). nstpcouple != 1 means that you simulate on an NVE manifold for a period of time (maybe with some T changes if nsttcouple != nstpcouple), and I'd suppose the longer that interval the bigger the chance of a build-up of pressure that P-R will then try to relieve by changing the box size. Larger nstlist and dt will exacerbate this, of course. I would recommend you experiment and see how far you can push things and keep statistics that still resemble those with small nstpcouple. Larger nstpcouple helps reduce the frequency with which global communication occurs, and that affects your simulation rate... life is complex! It would be nice if we were able to compute heuristics so that mdrun could anticipate such a problem and warn you, but off-hand that seems a tricky problem... The simulation runs with nstcalclr and nsttcouple set to 50 along with nstcalclr should have no effect - it works only with the group scheme, which does not work on GPUs. nstlist. Is nstlist the only setting that should be increased when utilising gpus? Yes, AFAIK. The point is that nstlist is the interval between neighbour searches, and (at the moment at least) that's only done on the CPU. The Verlet kernels cheerfully compute lots of zero-strength interactions outside the cutoff (by design), and depending on the relative performance of your hardware it is normally more efficient to bump nstlist up (and rlist accordingly, to provide a larger buffer for diffusion of particles) and compute more zeroes than it is to search for neighbours more often. Mark Thanks in advance, -Trayder P.S. The working mdp file: -- View this message in context: http://gromacs.5086.x6.nabble.com/Effect-of-pressure-coupling-frequency-on-gpu-simulations-tp5008439.html Sent from the GROMACS Users Forum mailing list archive at Nabble.com. -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
