<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>
Hi,<br>the center of mass of my channel is at the middle of the ion channel, and it is a symmetric system, so I suppose these results would be OK. Anyway, I will check the options you propose.<br>Thanks a lot for all your comments!!<br>Best wishes,<br>Rebeca.<br><br><div><div id="SkyDrivePlaceholder"></div>> Date: Thu, 31 May 2012 20:08:26 +0200<br>> From: schlesi@uni-mainz.de<br>> To: gmx-users@gromacs.org<br>> Subject: [gmx-users] Re: boundaries in PMF<br>> <br>> Where is the center of mass of reference group (MOL) located?<br>> <br>> It seems that the COM is near the middle of the ion channel. Since you <br>> use 'pull_geometry=distance', g_wham will look only for the distance <br>> between 'MOL' and 'Na' and that leads to problem.<br>> If the com of 'MOL' sits in the center of the channel (which is around <br>> 5nm long), one has 2.5nm in both directions. Since g_wham uses only the <br>> distance you get the PMF for half of the channel, but with the data of <br>> both parts.<br>> If the channel would be symmetric and the com of 'MOL' would lie exactly <br>> in the middle of the channel, this could be ok. But if even one of both <br>> assumptions is wrong, the results would be errorous.<br>> <br>> A better approach would be to use 'pull_geometry=direction', since the <br>> you define a vector along which the windows lie and do not have the <br>> problem that the distance is in both directions (along positive and <br>> negative vector) the same.<br>> Only problem could be that g_wham supports 'pull_geometry=direction' <br>> only from gromacs 4.5.x (don't know this, since instead of umbrella <br>> smapling i use thermodynamik integration, where one uses constraints <br>> (instead of restraints) and integrates the constraint-force; but the <br>> conceptual problem with 'distance/direction' is the same).<br>> <br>> Another approach (with 'pull_geometry=distance') would be to use a <br>> reference group which is just outside of the channel (like going some <br>> steps away from the channel, along the vector which goes through the <br>> channel). If there is only water, it would be bad, because then the <br>> reference group would be move away.<br>> But then on could use the entry and exit of the channel as a reference <br>> point for two simulations. In the case that the entry is the reference <br>> group, the PMF would be ill defined near the entry, but from the <br>> simulation with the exit as reference you would get the right PMF for <br>> the entry region and vice versa.<br>> <br>> Greetings<br>> Thomas<br>> <br>> <br>> Am 31.05.2012 19:39, schrieb gmx-users-request@gromacs.org:<br>> > Thanks a lot for your quick answer. The mdp file I have used is copied<br>> > below. What is strange is that when I look at the *gro files for the<br>> > different windows (100 windows in total), i. e: window 1: 8704Na Na56458<br>> > 5.134 5.085 5.824 window 50: 8704Na Na56458 5.053 5.081 3.459 window<br>> > 100: 8704Na Na56458 4.990 5.042 0.951 you can see that the z-coordinate<br>> > goes from 0.951 to 5.824 nm I should have a total distance in the x-axis<br>> > of about 5 nm, and however, the boundaries calculated by g_wham are<br>> > "Determined boundaries to 0.000035 to 2.603290 " Can you see anything in<br>> > the mdp file which is causing this behaviour...? Thanks again for your<br>> > help! MDP FILE USED: title = Umbrella pulling simulation define =<br>> > -DPOSRES_B define = ; Run parameters integrator = md dt = 0.002 tinit =<br>> > 0 nsteps = 500000 ; 1 ns nstcomm = 10 ; Output parameters nstxout = 5000<br>> > ; every 10 ps nstvout = 5000 nstfout = 5000 nstxtcout = 5000 ; every 10<br>> > ps nstenergy = 5000 ; Bond parameters constraint_algorithm = lincs<br>> > constraints = all-bonds continuation = yes ; Single-range cutoff scheme<br>> > nstlist = 5 ns_type = grid rlist = 1.4 rcoulomb = 1.4 rvdw = 1.4 ; PME<br>> > electrostatics parameters coulombtype = PME fourierspacing = 0.12<br>> > fourier_nx = 0 fourier_ny = 0 fourier_nz = 0 pme_order = 4 ewald_rtol =<br>> > 1e-5 optimize_fft = yes ;Berendsen temperature coupling is on Tcoupl =<br>> > V-rescale tau_t = 0.1 0.1 0.1 tc-grps = MOL dop WAT_Cl_Na ref_t = 300<br>> > 300 300 ; Pressure coupling Pcoupl = Parrinello-Rahman Pcoupltype =<br>> > Semiisotropic ; Time constant (ps), compressibility (1/bar) and<br>> > reference P (bar) tau-p = 1.0 1.0 compressibility = 4.6E-5 4.6E-5 ref-p<br>> > = 1.0 1.0 ; Generate velocities is off gen_vel = no ; Periodic boundary<br>> > conditions are on in all directions pbc = xyz ; Long-range dispersion<br>> > correction DispCorr = EnerPres ; Pull code pull = umbrella pull_geometry<br>> > = distance pull_dim = N N Y pull_start = yes pull_ngroups = 1<br>> > pull_group0 = MOL pull_group1 = Na pull_init1 = 0 pull_rate1 = 0.0<br>> > pull_k1 = 1000 ; kJ mol^-1 nm^-2 pull_nstxout = 1000 ; every 2 ps<br>> > pull_nstfout = 1000 ; every 2 ps<br>> >> > Date: Thu, 31 May 2012 13:25:26 -0400<br>> >> > From:jalemkul@vt.edu<br>> >> > To:gmx-users@gromacs.org<br>> >> > Subject: Re: [gmx-users] boundaries in PMF<br>> >> ><br>> >> ><br>> >> ><br>> >> > On 5/31/12 1:20 PM, Rebeca Garc?a Fandi?o wrote:<br>> >>> > > Hi,<br>> >>> > > I am trying to calculate a PMF for an ion along a channel. Everything went OK,<br>> >>> > > but when I used g_wham I got a profile with strange dimensions for the x-axis.<br>> >>> > > What are the boundaries g_wham is using for calculating the units of x-axis?<br>> >>> > ><br>> >> ><br>> >> > The values are based on the restraint distances along the reaction coordinate.<br>> >> ><br>> >>> > > I have used:<br>> >>> > > g_wham -it tpr-files.dat -if pullf-files.dat -o file_output.xvg -hist<br>> >>> > > file_histo_output.xvg -unit kCal -cycl yes<br>> >>> > ><br>> >>> > > (version 4.0.7)<br>> >>> > ><br>> >>> > > and the units I got in the x-axis are determined by the boundaries:<br>> >>> > ><br>> >>> > > "Determined boundaries to 0.000035 to 2.603290"<br>> >>> > ><br>> >>> > > Which are these units? nm?<br>> >>> > ><br>> >> ><br>> >> > Yes.<br>> >> ><br>> >>> > > The z coordinate for my ion explores at least 5 nm!!<br>> >>> > ><br>> >>> > > I am a bit confused about that.<br>> >>> > ><br>> >> ><br>> >> > The exact output depends on how you set up the umbrella sampling (in the .mdp<br>> >> > file). The range of values corresponds to whatever the distances are that are<br>> >> > sampled in the various windows. Perhaps there is a sign issue here? Do you<br>> >> > have some restraints that are at negative displacement and others at positive?<br>> >> > Did you set up the .mdp files properly to account for this behavior?<br>> >> ><br>> >> > -Justin<br>> <br>> -- <br>> gmx-users mailing list gmx-users@gromacs.org<br>> http://lists.gromacs.org/mailman/listinfo/gmx-users<br>> Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!<br>> Please don't post (un)subscribe requests to the list. Use the <br>> www interface or send it to gmx-users-request@gromacs.org.<br>> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists<br></div>                                            </div></body>
</html>