Re: [gmx-users] High Pressure with Martini system
On 25.04.2014 00:07, Ricardo O. S. Soares wrote: I ran an NVT equilibration of a 24mi CG Martini atoms and detected the formation of vacuum spots on the solvent. CG Martini has many sphere types, did you solvate your system with 24x10^6 MARTINI-W segments? MARTINI-W is is strongly association LJ-fluid. In large, inhomogeneous systems (non-optimal box solvation) it is expected to generate huge forces if either too-close or not close enough packed. However, the log file shows a very large negative pressure (about -4e+28) all the time, since the beginning. Despite the negative pressure, when I turn pressure coupling ON, the very first step causes the box to expand (!) too much and then the simulation crashes. there you are ... Things I tried, and failed: changed the tau_p to several values (1 to 500); changed the compressibility; changed the ref_p from 1 to values closer to the starting pressure; changed time-step from 0.02 up to 0.0005. Additional info: used the insane.py script to solvate the box with polarizable water; there's a large (frozen) bilayer at the center of the box, It, however is smaller than the box sides, to it doesn't interact with itself via PBC. I'm sure I could provide more info, so please ask me for the specifics and I'll reply. How large is the box, is it cubic? How many Lipids in the bilayer? What do you expect from the frozen bilayer if you pressure-couple your box dimensions? Be warned that the MARTINI-Lipids constitute bilayers which are very floppy and spontaneously fold into spheres (vesicles) if the aggregate has more than around 600 Molecules. BTW.: what hardware are you on if you expect to equilibrate a system of that size (if the 2.4x10^6 figure is correct?) Regards M. Disclaimer: I did simulate similar MARTINI systems of 1/10th the size of yours -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] High Pressure with Martini system
Hi Ricardo, Insanely built systems with polarizable watets may be tricky, especially if the system is large (not to say humongous). Maybe it's worth adding a stage using normal Martini water, let that equilibrate (NVT and NpT), and then add in the polarizable sites, letting the system equilibrate again (NVT, since the system size should pretty much match the equilibrium pressure density). Cheers, Tsjerk On Apr 25, 2014 8:06 AM, Mirco Wahab mirco.wa...@chemie.tu-freiberg.de wrote: On 25.04.2014 00:07, Ricardo O. S. Soares wrote: I ran an NVT equilibration of a 24mi CG Martini atoms and detected the formation of vacuum spots on the solvent. CG Martini has many sphere types, did you solvate your system with 24x10^6 MARTINI-W segments? MARTINI-W is is strongly association LJ-fluid. In large, inhomogeneous systems (non-optimal box solvation) it is expected to generate huge forces if either too-close or not close enough packed. However, the log file shows a very large negative pressure (about -4e+28) all the time, since the beginning. Despite the negative pressure, when I turn pressure coupling ON, the very first step causes the box to expand (!) too much and then the simulation crashes. there you are ... Things I tried, and failed: changed the tau_p to several values (1 to 500); changed the compressibility; changed the ref_p from 1 to values closer to the starting pressure; changed time-step from 0.02 up to 0.0005. Additional info: used the insane.py script to solvate the box with polarizable water; there's a large (frozen) bilayer at the center of the box, It, however is smaller than the box sides, to it doesn't interact with itself via PBC. I'm sure I could provide more info, so please ask me for the specifics and I'll reply. How large is the box, is it cubic? How many Lipids in the bilayer? What do you expect from the frozen bilayer if you pressure-couple your box dimensions? Be warned that the MARTINI-Lipids constitute bilayers which are very floppy and spontaneously fold into spheres (vesicles) if the aggregate has more than around 600 Molecules. BTW.: what hardware are you on if you expect to equilibrate a system of that size (if the 2.4x10^6 figure is correct?) Regards M. Disclaimer: I did simulate similar MARTINI systems of 1/10th the size of yours -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] NPT simulation stage
On 4/25/14, 2:36 AM, Mahboobeh Eslami wrote: hi GMX users please help me i want to simulate protein-ligand complex by gromacs4.6.5 when i repeat NPT simulation stage many times, simulation results are very different for the pressure average.I sincerely thank you for your guidance Pressure is probably the most ill-defined quantity in MD simulations. It can fluctuate dramatically. Without the actual numbers, it's hard to comment, but issues related to pressure have been discussed extensively over the list in the past and there is useful information in http://www.gromacs.org/Documentation/Terminology/Pressure. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA 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 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] MD simullations with Morse potential
Dear All, I am trying to do a simulations with morse potential for bonds, I can do the minimization of the system, but I can not do the MD optimization without lincs constraints on bonds since using Morse potential with constraints is useless. The error was shown as Segmentation fault (core dumped) , I did not found some other information in the log file. The system was well minimized before doing md optimization. Here is my mdp file: ## title = OPLS Lysozyme NPT equilibration ;define = -DPOSRES ; position restrain the protein ; Run parameters integrator = md ; leap-frog integrator nsteps = 10; 2 * 5 = 100 ps dt = 0.0005; 2 fs ; Output control nstxout = 2500; save coordinates every 0.2 ps nstvout= 2500; save velocities every 0.2 ps nstenergy= 2500 ; save energies every 0.2 ps nstlog = 2500 ; update log file every 0.2 ps ; Bond parameters continuation= no ; Restarting after NVT ;constraint_algorithm = lincs ; holonomic constraints ;constraints = all-bonds ; all bonds (even heavy atom-H bonds) constrained ;lincs_iter= 1 ; accuracy of LINCS ;lincs_order = 4 ; also related to accuracy ; Neighborsearching ns_type = grid ; search neighboring grid cells nstlist = 5 ; 10 fs rlist= 1.2 ; short-range neighborlist cutoff (in nm) rcoulomb= 1.2 ; short-range electrostatic cutoff (in nm) rvdw = 1.2 ; short-range van der Waals cutoff (in nm) ; Electrostatics coulombtype = PME; Particle Mesh Ewald for long-range electrostatics pme_order = 4 ; cubic interpolation fourierspacing = 0.16 ; grid spacing for FFT ; Temperature coupling is on tcoupl = V-rescale ; modified Berendsen thermostat tc-grps = Protein Non-Protein ; two coupling groups - more accurate tau_t= 0.1 0.1 ; time constant, in ps ref_t = 300 300 ; reference temperature, one for each group, in K ; Pressure coupling is on pcoupl = Parrinello-Rahman; Pressure coupling on in NPT pcoupltype= isotropic ; uniform scaling of box vectors tau_p = 2.0 ; time constant, in ps ref_p= 1.0 ; reference pressure, in bar compressibility = 4.5e-5 ; isothermal compressibility of water, bar^-1 refcoord_scaling = com ; Periodic boundary conditions pbc = xyz ; 3-D PBC ; Dispersion correction DispCorr = EnerPres ; account for cut-off vdW scheme ; Velocity generation gen_vel= yes ; Velocity generation is off gen_temp= 300 ; temperature for Maxwell distribution gen_seed= 173529 ; generate a random seed morse = yes Is there some special rule for md simulations with morse potential? But I did not found it in the manual. Could someone give some suggestions? Any response will be highly appreciated! All the best, Qinghua -- View this message in context: http://gromacs.5086.x6.nabble.com/MD-simullations-with-Morse-potential-tp5016062.html Sent from the GROMACS Users Forum mailing list archive at Nabble.com. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] compilation issues with GCC/clang/CUDA on OSX
Hi!
Note that the warnings are only suggesting that your build may give suboptimal
performance. (as in having no compatible GPU or non-openmp enabled compiler).
In my experience, you can actually ignore the first one about not finding GPUs.
That never worked for me either (on mac os x).
Did you try proceed to 'make' anything?
The openmp-related warning suggest to me that either cmake is confused about
what compiler to use (clear out the build dir and try again?) or your gcc
installation is somehow broken.
Happy compiling!
Anders
On 24 Apr 2014, at 21:41, Stephen N. Floor fl...@berkeley.edu wrote:
Hello -
I am trying to compile gromacs 4.6.5 with CUDA support for a new
graphics card on OSX 10.8.5 with CUDA 6.0.37 using gcc 4.8.2 from fink
(gcc-4 and g++-4) for OpenMP support and clang for CUDA as recommended
previously.
I am starting cmake like this:
export CC=/sw/bin/gcc-4
export CXX=/sw/bin/g++-4
cmake -DGMX_GPU=ON -DCMAKE_C_COMPILER=${CC}
-DCMAKE_CXX_COMPILER=${CXX} -DCUDA_HOST_COMPILER=/usr/bin/clang
-DCUDA_PROPAGATE_HOST_FLAGS=OFF ..
I am having two issues:
1) cmake does not detect any video cards:
-- Looking for NVIDIA GPUs present in the system
-- Could not detect NVIDIA GPUs
-- Found CUDA: /usr/local/cuda (found suitable version 6.0, minimum
required is 3.2)
While I can likely continue with compilation, this seems to suggest an
underlying issue that may rear its head later.
2) OpenMP is not detected
{ snipped many openmp flag test results that are all failures }
CMake Warning at CMakeLists.txt:273 (message):
The compiler you are using does not support OpenMP parallelism. This might
hurt your performance a lot, in particular with GPUs. Try using a more
recent version, or a different compiler. For now, we are proceeding by
turning off OpenMP.
compiler info:
[snf@waipio gromacs/gromacs-4.6.5/build-cmake-gpu]$ gcc-4 --version -v
Using built-in specs.
COLLECT_GCC=gcc-4
COLLECT_LTO_WRAPPER=/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.5.0/4.8.2/lto-wrapper
gcc-4 (GCC) 4.8.2
Copyright (C) 2013 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Target: x86_64-apple-darwin12.5.0
Configured with: ../gcc-4.8.2/configure --prefix=/sw
--prefix=/sw/lib/gcc4.8 --mandir=/sw/share/man
--infodir=/sw/lib/gcc4.8/info
--enable-languages=c,c++,fortran,lto,objc,obj-c++,java --with-gmp=/sw
--with-libiconv-prefix=/sw --with-isl=/sw --with-cloog=/sw
--with-mpc=/sw --with-system-zlib --x-includes=/usr/X11R6/include
--x-libraries=/usr/X11R6/lib --program-suffix=-fsf-4.8
Thread model: posix
gcc version 4.8.2 (GCC)
COLLECT_GCC_OPTIONS='-mmacosx-version-min=10.8.5' '--version' '-v'
'-mtune=core2'
/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.5.0/4.8.2/cc1 -quiet
-v -D__DYNAMIC__ help-dummy -fPIC -quiet -dumpbase help-dummy
-mmacosx-version-min=10.8.5 -mtune=core2 -auxbase help-dummy -version
--version -o /var/folders/bj/fl_w4q152yx6cg787vx5c12hgq/T//ccQjwwGf.s
GNU C (GCC) version 4.8.2 (x86_64-apple-darwin12.5.0)
compiled by GNU C version 4.8.2, GMP version 5.1.3, MPFR
version 3.1.2, MPC version 1.0.2
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
COLLECT_GCC_OPTIONS='-mmacosx-version-min=10.8.5' '--version' '-v'
'-mtune=core2'
as -arch x86_64 -force_cpusubtype_ALL --version -o
Thanks for any insight!
-Stephen
--
Stephen N. Floor
HHWF Postdoctoral Fellow, Doudna Group
http://doudna.berkeley.edu
--
Gromacs Users mailing list
* Please search the archive at
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a
mail to gmx-users-requ...@gromacs.org.
--
Gromacs Users mailing list
* Please search the archive at
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a
mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] MD simullations with Morse potential
On 4/25/14, 10:00 AM, fantasticqhl wrote: Dear All, I am trying to do a simulations with morse potential for bonds, I can do the minimization of the system, but I can not do the MD optimization without lincs constraints on bonds since using Morse potential with constraints is useless. The error was shown as Segmentation fault (core dumped) , I did not found some other information in the log file. The system was well minimized before doing md optimization. Here is my mdp file: ## title = OPLS Lysozyme NPT equilibration ;define = -DPOSRES ; position restrain the protein ; Run parameters integrator = md ; leap-frog integrator nsteps = 10; 2 * 5 = 100 ps dt = 0.0005; 2 fs ; Output control nstxout = 2500; save coordinates every 0.2 ps nstvout= 2500; save velocities every 0.2 ps nstenergy= 2500 ; save energies every 0.2 ps nstlog = 2500 ; update log file every 0.2 ps ; Bond parameters continuation= no ; Restarting after NVT ;constraint_algorithm = lincs ; holonomic constraints ;constraints = all-bonds ; all bonds (even heavy atom-H bonds) constrained ;lincs_iter= 1 ; accuracy of LINCS ;lincs_order = 4 ; also related to accuracy ; Neighborsearching ns_type = grid ; search neighboring grid cells nstlist = 5 ; 10 fs rlist= 1.2 ; short-range neighborlist cutoff (in nm) rcoulomb= 1.2 ; short-range electrostatic cutoff (in nm) rvdw = 1.2 ; short-range van der Waals cutoff (in nm) ; Electrostatics coulombtype = PME; Particle Mesh Ewald for long-range electrostatics pme_order = 4 ; cubic interpolation fourierspacing = 0.16 ; grid spacing for FFT ; Temperature coupling is on tcoupl = V-rescale ; modified Berendsen thermostat tc-grps = Protein Non-Protein ; two coupling groups - more accurate tau_t= 0.1 0.1 ; time constant, in ps ref_t = 300 300 ; reference temperature, one for each group, in K ; Pressure coupling is on pcoupl = Parrinello-Rahman; Pressure coupling on in NPT pcoupltype= isotropic ; uniform scaling of box vectors tau_p = 2.0 ; time constant, in ps ref_p= 1.0 ; reference pressure, in bar compressibility = 4.5e-5 ; isothermal compressibility of water, bar^-1 refcoord_scaling = com ; Periodic boundary conditions pbc = xyz ; 3-D PBC ; Dispersion correction DispCorr = EnerPres ; account for cut-off vdW scheme ; Velocity generation gen_vel= yes ; Velocity generation is off gen_temp= 300 ; temperature for Maxwell distribution gen_seed= 173529 ; generate a random seed morse = yes Is there some special rule for md simulations with morse potential? But I did not found it in the manual. Could someone give some suggestions? Any response will be highly appreciated! The first problem I suspect is the use of the Parrinello-Rahman barostat while generating velocities. That combination often results in instability for any system. Try NVT equilibration before moving to NPT. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA 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 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] compilation issues with GCC/clang/CUDA on OSX
Hi and thanks -
I did end up ignoring the warning about no compatible GPUs. The non-openmp
compiler went away once I finally convinced cmake to actually use gcc-4.8.2
instead of clang - embarrassingly, it seems I was getting those errors because
I was running cmake repeatedly without wiping the directory in between. Making
a fresh build dir and running cmake as listed below does recognize openmp
compatibility via gcc, and CUDA things build fine using clang. After ignoring
the GPU warning it builds fine and does detect a GPU when I perform
simulations.
However, I ran into the Malformed Mach-o file issue that was mentioned by a
previous poster. After make finishes, I run make install and receive a huge
number of errors like this:
CMake Error:
Error evaluating generator expression:
$LINK_ONLY:-Wl,-rpath
$LINK_ONLY expression requires exactly one parameter.
CMake Warning (dev) in src/gmxlib/gpu_utils/CMakeLists.txt:
Policy CMP0022 is not set: INTERFACE_LINK_LIBRARIES defines the link
interface. Run cmake --help-policy CMP0022 for policy details. Use the
cmake_policy command to set the policy and suppress this warning.
Static library target gpu_utils has a INTERFACE_LINK_LIBRARIES property.
This should be preferred as the source of the link interface for this
library. Ignoring the property and using the link implementation as the
link interface instead.
This warning is for project developers. Use -Wno-dev to suppress it.
If I repeat make install then these errors do not reappear and it does install
to the specified prefix. However, executables that are installed in this
manner are broken:
[snf@waipio ]$ nice -n 19 mdrun -deffnm npt-10ps
nice: mdrun: Malformed Mach-o file
A Mach-o file is an OSX format executable, similar to ELF. This error occurs
for any gromacs package executable that I have tried. However, if I go into
the freshly compiled src tree I can run mdrun or anything else just fine, which
suggests to me that something in the make install process is garping things up.
Anyway, I did find a workaround for this by just using the executables in the
src tree of the build directory, but thought I should share.
Thanks,
-Stephen
On Apr 25, 2014, at 7:43 AM, Anders Gabrielsson and...@kth.se wrote:
Hi!
Note that the warnings are only suggesting that your build may give
suboptimal performance. (as in having no compatible GPU or non-openmp enabled
compiler).
In my experience, you can actually ignore the first one about not finding
GPUs. That never worked for me either (on mac os x).
Did you try proceed to 'make' anything?
The openmp-related warning suggest to me that either cmake is confused about
what compiler to use (clear out the build dir and try again?) or your gcc
installation is somehow broken.
Happy compiling!
Anders
On 24 Apr 2014, at 21:41, Stephen N. Floor fl...@berkeley.edu wrote:
Hello -
I am trying to compile gromacs 4.6.5 with CUDA support for a new
graphics card on OSX 10.8.5 with CUDA 6.0.37 using gcc 4.8.2 from fink
(gcc-4 and g++-4) for OpenMP support and clang for CUDA as recommended
previously.
I am starting cmake like this:
export CC=/sw/bin/gcc-4
export CXX=/sw/bin/g++-4
cmake -DGMX_GPU=ON -DCMAKE_C_COMPILER=${CC}
-DCMAKE_CXX_COMPILER=${CXX} -DCUDA_HOST_COMPILER=/usr/bin/clang
-DCUDA_PROPAGATE_HOST_FLAGS=OFF ..
I am having two issues:
1) cmake does not detect any video cards:
-- Looking for NVIDIA GPUs present in the system
-- Could not detect NVIDIA GPUs
-- Found CUDA: /usr/local/cuda (found suitable version 6.0, minimum
required is 3.2)
While I can likely continue with compilation, this seems to suggest an
underlying issue that may rear its head later.
2) OpenMP is not detected
{ snipped many openmp flag test results that are all failures }
CMake Warning at CMakeLists.txt:273 (message):
The compiler you are using does not support OpenMP parallelism. This might
hurt your performance a lot, in particular with GPUs. Try using a more
recent version, or a different compiler. For now, we are proceeding by
turning off OpenMP.
compiler info:
[snf@waipio gromacs/gromacs-4.6.5/build-cmake-gpu]$ gcc-4 --version -v
Using built-in specs.
COLLECT_GCC=gcc-4
COLLECT_LTO_WRAPPER=/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.5.0/4.8.2/lto-wrapper
gcc-4 (GCC) 4.8.2
Copyright (C) 2013 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Target: x86_64-apple-darwin12.5.0
Configured with: ../gcc-4.8.2/configure --prefix=/sw
--prefix=/sw/lib/gcc4.8 --mandir=/sw/share/man
--infodir=/sw/lib/gcc4.8/info
--enable-languages=c,c++,fortran,lto,objc,obj-c++,java --with-gmp=/sw
--with-libiconv-prefix=/sw --with-isl=/sw --with-cloog=/sw
--with-mpc=/sw --with-system-zlib --x-includes=/usr/X11R6/include
Re: [gmx-users] High Pressure with Martini system
Hello Mirco, thanks for your reply. Please check my notes below. Ricardo. De: Mirco Wahab mirco.wa...@chemie.tu-freiberg.de Para: gromacs org gmx-users gromacs.org_gmx-users@maillist.sys.kth.se Enviadas: Sexta-feira, 25 de Abril de 2014 3:03:18 Assunto: Re: [gmx-users] High Pressure with Martini system On 25.04.2014 00:07, Ricardo O. S. Soares wrote: I ran an NVT equilibration of a 24mi CG Martini atoms and detected the formation of vacuum spots on the solvent. CG Martini has many sphere types, did you solvate your system with 24x10^6 MARTINI-W segments? The whole system has about 24x10⁶ atoms/beads. Since I used PW (polarizable water, which has 3 'atoms' -- WM, WP, W), I have about 7.5x10^6 molecules, and therefore about 22.5 x 10^6 solvent atoms. MARTINI-W is is strongly association LJ-fluid. In large, inhomogeneous systems (non-optimal box solvation) it is expected to generate huge forces if either too-close or not close enough packed. However, the log file shows a very large negative pressure (about -4e+28) all the time, since the beginning. Despite the negative pressure, when I turn pressure coupling ON, the very first step causes the box to expand (!) too much and then the simulation crashes. there you are ... I'm just failing to understand why a negative pressure would cause the box to expand, instead of contract... Things I tried, and failed: changed the tau_p to several values (1 to 500); changed the compressibility; changed the ref_p from 1 to values closer to the starting pressure; changed time-step from 0.02 up to 0.0005. Additional info: used the insane.py script to solvate the box with polarizable water; there's a large (frozen) bilayer at the center of the box, It, however is smaller than the box sides, to it doesn't interact with itself via PBC. I'm sure I could provide more info, so please ask me for the specifics and I'll reply. How large is the box, is it cubic? How many Lipids in the bilayer? The box dimensions are 153.32965 102.32965 62.0, and I have XXX lipids. What do you expect from the frozen bilayer if you pressure-couple your box dimensions? The bilayer is smaller than the XY dimensions of the box, so I'd expect it to remain frozen until the desired pressure and solvation is reached and then I'd remove the freezing code and probably add some weak restrains in the Z axis, for further equilibration. Be warned that the MARTINI-Lipids constitute bilayers which are very floppy and spontaneously fold into spheres (vesicles) if the aggregate has more than around 600 Molecules. That interest me, do you have a reference for this information? thanks. BTW.: what hardware are you on if you expect to equilibrate a system of that size (if the 2.4x10^6 figure is correct?) For this system I'm using a BlueGene/P an SGI Altix XE 1300 . Regards M. Disclaimer: I did simulate similar MARTINI systems of 1/10th the size of yours I did simulated a smaller (500K atoms) similar system before with no problems, but then again, 24x10^6 isn't 500K ;) -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] High Pressure with Martini system
Hello Tsjerk, You're right, when insanely solvating I sometimes put the water too close or too far. Do you know if 0.47 is a good value for the -sold parameter? From my experience with Martini-W beads and a small bilayer, I usually find that the system tends to freeze. However, I think that this happens after the pressure reaches a reasonable value, so I could in this moment use the triple-w.py script to replace W for PW, right? And then, like you said, perform NVT and NpT equilibration. .. Thanks for your input, Best, Ricardo. --- Biological Chemistry and Physics Faculty of Pharmaceutical Sciences at Ribeirão Preto University of São Paulo - Brazil - Mensagem original - De: Tsjerk Wassenaar tsje...@gmail.com Para: Discussion list for GROMACS users gmx-us...@gromacs.org Enviadas: Sexta-feira, 25 de Abril de 2014 3:31:52 Assunto: Re: [gmx-users] High Pressure with Martini system Hi Ricardo, Insanely built systems with polarizable watets may be tricky, especially if the system is large (not to say humongous). Maybe it's worth adding a stage using normal Martini water, let that equilibrate (NVT and NpT), and then add in the polarizable sites, letting the system equilibrate again (NVT, since the system size should pretty much match the equilibrium pressure density). Cheers, Tsjerk On Apr 25, 2014 8:06 AM, Mirco Wahab mirco.wa...@chemie.tu-freiberg.de wrote: On 25.04.2014 00:07, Ricardo O. S. Soares wrote: I ran an NVT equilibration of a 24mi CG Martini atoms and detected the formation of vacuum spots on the solvent. CG Martini has many sphere types, did you solvate your system with 24x10^6 MARTINI-W segments? MARTINI-W is is strongly association LJ-fluid. In large, inhomogeneous systems (non-optimal box solvation) it is expected to generate huge forces if either too-close or not close enough packed. However, the log file shows a very large negative pressure (about -4e+28) all the time, since the beginning. Despite the negative pressure, when I turn pressure coupling ON, the very first step causes the box to expand (!) too much and then the simulation crashes. there you are ... Things I tried, and failed: changed the tau_p to several values (1 to 500); changed the compressibility; changed the ref_p from 1 to values closer to the starting pressure; changed time-step from 0.02 up to 0.0005. Additional info: used the insane.py script to solvate the box with polarizable water; there's a large (frozen) bilayer at the center of the box, It, however is smaller than the box sides, to it doesn't interact with itself via PBC. I'm sure I could provide more info, so please ask me for the specifics and I'll reply. How large is the box, is it cubic? How many Lipids in the bilayer? What do you expect from the frozen bilayer if you pressure-couple your box dimensions? Be warned that the MARTINI-Lipids constitute bilayers which are very floppy and spontaneously fold into spheres (vesicles) if the aggregate has more than around 600 Molecules. BTW.: what hardware are you on if you expect to equilibrate a system of that size (if the 2.4x10^6 figure is correct?) Regards M. Disclaimer: I did simulate similar MARTINI systems of 1/10th the size of yours -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] High Pressure with Martini system
Hi Ricardo, I think 0.47 should be fine. For normal W you can also increase -solr to get a bit less ordering, although you need to be careful with large systems. It should also be possible to add a certain number of antifreeze particles, but I think you should be fine converting to PW after a short equilibration with W. Cheers, Tsjerk On Apr 25, 2014 8:54 PM, Ricardo O. S. Soares rsoa...@fcfrp.usp.br wrote: Hello Tsjerk, You're right, when insanely solvating I sometimes put the water too close or too far. Do you know if 0.47 is a good value for the -sold parameter? From my experience with Martini-W beads and a small bilayer, I usually find that the system tends to freeze. However, I think that this happens after the pressure reaches a reasonable value, so I could in this moment use the triple-w.py script to replace W for PW, right? And then, like you said, perform NVT and NpT equilibration. .. Thanks for your input, Best, Ricardo. --- Biological Chemistry and Physics Faculty of Pharmaceutical Sciences at Ribeirão Preto University of São Paulo - Brazil - Mensagem original - De: Tsjerk Wassenaar tsje...@gmail.com Para: Discussion list for GROMACS users gmx-us...@gromacs.org Enviadas: Sexta-feira, 25 de Abril de 2014 3:31:52 Assunto: Re: [gmx-users] High Pressure with Martini system Hi Ricardo, Insanely built systems with polarizable watets may be tricky, especially if the system is large (not to say humongous). Maybe it's worth adding a stage using normal Martini water, let that equilibrate (NVT and NpT), and then add in the polarizable sites, letting the system equilibrate again (NVT, since the system size should pretty much match the equilibrium pressure density). Cheers, Tsjerk On Apr 25, 2014 8:06 AM, Mirco Wahab mirco.wa...@chemie.tu-freiberg.de wrote: On 25.04.2014 00:07, Ricardo O. S. Soares wrote: I ran an NVT equilibration of a 24mi CG Martini atoms and detected the formation of vacuum spots on the solvent. CG Martini has many sphere types, did you solvate your system with 24x10^6 MARTINI-W segments? MARTINI-W is is strongly association LJ-fluid. In large, inhomogeneous systems (non-optimal box solvation) it is expected to generate huge forces if either too-close or not close enough packed. However, the log file shows a very large negative pressure (about -4e+28) all the time, since the beginning. Despite the negative pressure, when I turn pressure coupling ON, the very first step causes the box to expand (!) too much and then the simulation crashes. there you are ... Things I tried, and failed: changed the tau_p to several values (1 to 500); changed the compressibility; changed the ref_p from 1 to values closer to the starting pressure; changed time-step from 0.02 up to 0.0005. Additional info: used the insane.py script to solvate the box with polarizable water; there's a large (frozen) bilayer at the center of the box, It, however is smaller than the box sides, to it doesn't interact with itself via PBC. I'm sure I could provide more info, so please ask me for the specifics and I'll reply. How large is the box, is it cubic? How many Lipids in the bilayer? What do you expect from the frozen bilayer if you pressure-couple your box dimensions? Be warned that the MARTINI-Lipids constitute bilayers which are very floppy and spontaneously fold into spheres (vesicles) if the aggregate has more than around 600 Molecules. BTW.: what hardware are you on if you expect to equilibrate a system of that size (if the 2.4x10^6 figure is correct?) Regards M. Disclaimer: I did simulate similar MARTINI systems of 1/10th the size of yours -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. --
Re: [gmx-users] MD simullations with Morse potential
Dear Justin, Thanks very much for your reply! I guess the use of the Parrinello-Rahman barostat should not be a problem because the optimization ran successfully if I turned off the morse potential. Actually, I have already done the nvt optimization without morse potential before I moved to npt optimization with morse potential, the error was the same for nvt with morse potential. I just do not know why the morse potential was not working. Do you have some other ideas which may result in the problem? Thanks very much! All the best, Qinghua -- View this message in context: http://gromacs.5086.x6.nabble.com/MD-simullations-with-Morse-potential-tp5016062p5016072.html Sent from the GROMACS Users Forum mailing list archive at Nabble.com. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] MD simullations with Morse potential
On 4/25/14, 5:37 PM, fantasticqhl wrote: Dear Justin, Thanks very much for your reply! I guess the use of the Parrinello-Rahman barostat should not be a problem because the optimization ran successfully if I turned off the morse potential. Actually, I have already done the nvt optimization without morse potential before I moved to npt optimization with morse potential, the error was the same for nvt with morse potential. I just do not know why the morse potential was not working. Do you have some other ideas which may result in the problem? Thanks very much! Well, Morse potentials aren't the functional form with which any of the biomolecular force fields in Gromacs were parametrized, so either (1) there's a bug that needs to be fixed or (2) you're doing something incompatible and the model physics is breaking down. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA 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 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] MD simullations with Morse potential
Hi Qinghua, Can you run NVT equilibration? Does it work with nstlist set to 1? Cheers, Tsjerk On Apr 26, 2014 12:29 AM, Justin Lemkul jalem...@vt.edu wrote: On 4/25/14, 5:37 PM, fantasticqhl wrote: Dear Justin, Thanks very much for your reply! I guess the use of the Parrinello-Rahman barostat should not be a problem because the optimization ran successfully if I turned off the morse potential. Actually, I have already done the nvt optimization without morse potential before I moved to npt optimization with morse potential, the error was the same for nvt with morse potential. I just do not know why the morse potential was not working. Do you have some other ideas which may result in the problem? Thanks very much! Well, Morse potentials aren't the functional form with which any of the biomolecular force fields in Gromacs were parametrized, so either (1) there's a bug that needs to be fixed or (2) you're doing something incompatible and the model physics is breaking down. -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA 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 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/ Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
