<html>
<head>
<style>
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Verdana
}
</style>
</head>
<body class='hmmessage'>
Hi,<br><br>I just found out that I introduced a bug in 4.0.3, which could cause<br>the pull code to crash or give wrong results when running single processor.<br>In parallel it is correct.<br>I committed fixed for 4.0.4.<br><br>If you run in parallel things should be correct,<br>or change in src/kernel/md.c:<br> if (ir->pull) {<br> dd_make_local_pull_groups(NULL,ir->pull,mdatoms);<br>to<br> if (ir->pull && PAR(cr)) {<br> dd_make_local_pull_groups(NULL,ir->pull,mdatoms);<br><br>Berk<br><br>> Date: Fri, 23 Jan 2009 12:14:43 -0500<br>> From: fiedler@umich.edu<br>> To: gmx-users@gromacs.org<br>> Subject: Re: [gmx-users] Transition difficulties: version 3.3.3 to 4.0.3 regarding pull_geometry=distance<br>> <br>> Hi,<br>> <br>> Thank you Berk and Chris for the suggestions.<br>> <br>> To address the possibility that this issue is related to periodic <br>> boundaries, I used two approaches:<br>> 1. The pull group of interest (permeant) was centered in the x-y plane <br>> of the box using Chris' approach. I then used the genconf utility to <br>> replicate my lipid box to a 9x9 grid in the x-y plane and removed all <br>> but the center box. This generated the coordinates for a bilayer system <br>> with all lipid molecules inside a box and intact. The discrepancy <br>> between the grompp (version 4.0.3) output and distances as calculated by <br>> g_traj (version 4.0.3) persist, 2.667 vs. 0.3996 nm. <br>> 2. I constructed a three atom system containing 2 reference atoms of <br>> type A, and a "pull" atom of type B. Proper output from grompp was <br>> observed for all coordinates of both the reference and pulled atoms, <br>> include coordinates for atoms moved outside the box in the x-y plane. <br>> The coordinate, topology, and run control parameter file are given below.<br>> <br>> If there are additional suggestions, I would be greatly appreciative.<br>> <br>> Thank you,<br>> <br>> Steve Fiedler<br>> <br>> -----------------<br>> conf.gro<br>> Three atoms<br>> 3<br>> 1AAA A 1 1.500 1.500 1.000<br>> 2AAA A 2 0.500 1.500 1.000<br>> 3BBB B 3 -1.500 1.500 1.700<br>> 3.00000 3.00000 3.00000<br>> -----------------<br>> index.ndx<br>> [ System ]<br>> 1 2 3<br>> [ Ref ]<br>> 1 2<br>> [ Pulled ]<br>> 3<br>> -----------------<br>> grompp.mdp<br>> title = ThreeAtoms<br>> integrator = md<br>> dt = 0.001<br>> nsteps = 1<br>> ns_type = grid<br>> pbc = xyz<br>> coulombtype = shift<br>> rlist = 1.4<br>> rcoulomb = 1.4<br>> rvdw = 1.4<br>> tcoupl = no<br>> pcoupl = no<br>> constraint_algorithm = shake<br>> shake_tol = 1e-4<br>> gen-vel = no<br>> gen-temp = 0<br>> <br>> nstxout = 1<br>> nstvout = 0<br>> nstfout = 0<br>> <br>> pull = umbrella<br>> pull_geometry = distance<br>> pull_dim = N N Y<br>> pull_start = no<br>> pull_init1 = 0.7<br>> pull_group0 = Ref<br>> pull_group1 = Pulled<br>> pull_k1 = 10000<br>> -----------------<br>> topology.top<br>> ; topology for two partially charged atoms<br>> <br>> [ defaults ]<br>> ; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ<br>> 1 3 yes 0.125 0.5<br>> <br>> [ atomtypes ]<br>> ;name mass charge ptype sig eps<br>> A 1000.0000 0.000 A 0.50000 9.90000<br>> B 9.0000 0.000 A 0.30000 9.00000<br>> <br>> [ nonbond_params ]<br>> ; i j func sig eps<br>> <br>> [ moleculetype ]<br>> AAAA 1<br>> <br>> [ atoms ]<br>> ; nr type resnr residue atom cgnr charge mass<br>> 1 A 1 AAA A 1 0.000 1000.0000<br>> <br>> [ moleculetype ]<br>> BBBB 1<br>> <br>> [ atoms ]<br>> ; nr type resnr residue atom cgnr charge mass<br>> 1 B 1 BBB B 1 0.000 9.00<br>> <br>> [ system ]<br>> ; name<br>> Three atoms<br>> <br>> [ molecules ]<br>> ; name number<br>> AAAA 2<br>> BBBB 1<br>> <br>> <br>> <br>> <br>> <br>> Chris Neale wrote:<br>> > I just checked similar simulations of mine and Berk's suggestion <br>> > accounts for similar discrepancies that I notice on a quick evaluation <br>> > where g_traj and g_dist fail to give me the same distance as I obtain <br>> > from the pull pos.xvg file. As Berk suggests, once I first trjconv <br>> > -center -pbc mol -ur compact (giving an appropriate residue for <br>> > centering that puts all relevant pulled atoms in the same box) then <br>> > g_traj and g_dist both give me the exact same answer as I calculate <br>> > based on pull pos.xvg. Chris -- original message -- Hi, There could be <br>> > a problem with periodic boundary conditions. Do you have multiple <br>> > molecules in a pull group, or broken molecules? In that case the COM <br>> > position of 3.3.3 and g_traj are both incorrect. The pull code in 4.0 <br>> > grompp and mdrun are (as far as I know) always correct. Berk<br>> ><br>> >> > Date: Thu, 22 Jan 2009 13:22:24 -0500<br>> >> > From: fiedler@umich.edu<br>> >> > To: gmx-users@gromacs.org<br>> >> > Subject: [gmx-users] Transition difficulties: version 3.3.3 to <br>> >> 4.0.3 regarding pull_geometry=distance<br>> >> > > Dear all,<br>> >> > > I have encountered an odd behavior with use of the "pull_geometry <br>> >> = > distance" option of the pull code, upon transitioning from <br>> >> Gromacs > version 3.3.3 to version 4.0.3. It appears to be related <br>> >> to the center > of mass distances of the two pull groups, which has <br>> >> an effect of > abruptly displacing the coordinates of the less <br>> >> massive group. A > diagnostic is a discrepancy between the distances <br>> >> between the pull > groups from the preprocessor output in version <br>> >> 4.0.3, and the distance > between the groups as calculated using the <br>> >> difference of the groups' > centers of mass from the g_traj utility. <br>> >> For example, using the > coordinates of a system previously <br>> >> equilibrated with the constraint > force approach from version 3.3.3, <br>> >> the grompp output from version 4.0.3 is:<br>> >> > > Pull group natoms pbc atom distance at start reference at t=0<br>> >> > 0 2672 1336 > 1 60 6818 <br>> >> 2.673 0.400 > > Using g_traj (4.0.3 version), the <br>> >> difference of the distance between the > center of masses of the two <br>> >> groups is: 0.39911 nm versus the 2.673 value > from above. > > This <br>> >> issue does not exist in previous versions of Gromacs including > <br>> >> version 3.3.3. In version 4.0.3, this behavior occurs for both > <br>> >> pull=umbrella and pull = constraint, on 32 and 64 bit architecture > <br>> >> systems, and in both single and double precision calculations. A <br>> >> test > of a two atom system determined that the pull_start option was <br>> >> not > appropriate. The pull options used in the mdp file are listed <br>> >> below, as > well as the contents of the ppa file which has worked <br>> >> previously. > > Suggestions would be appreciated,<br>> > _______________________________________________<br>> > gmx-users mailing list gmx-users@gromacs.org<br>> > http://www.gromacs.org/mailman/listinfo/gmx-users<br>> > Please search the archive at http://www.gromacs.org/search before <br>> > posting!<br>> > Please don't post (un)subscribe requests to the list. Use the www <br>> > interface or send it to gmx-users-request@gromacs.org.<br>> > Can't post? Read http://www.gromacs.org/mailing_lists/users.php<br>> ><br>> ><br>> <br>> <br>> -- <br>> Steve Fiedler, Ph.D.<br>> Research Fellow<br>> Department of Mechanical Engineering<br>> The University of Michigan<br>> 2024 G.G. Brown<br>> 2350 Hayward St.<br>> Ann Arbor, MI 48109-2125<br>> <br>> <br>> _______________________________________________<br>> gmx-users mailing list gmx-users@gromacs.org<br>> http://www.gromacs.org/mailman/listinfo/gmx-users<br>> Please search the archive at http://www.gromacs.org/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/mailing_lists/users.php<br><br /><hr />Express yourself instantly with MSN Messenger! <a href='http://clk.atdmt.com/AVE/go/onm00200471ave/direct/01/' target='_new'>MSN Messenger</a></body>
</html>