Dear Mark and Arthur, thank you for your reply.<br>I did what you suggest me and almost everything worked.<br>I created a topology with the two molecules without bonded parameters<br>between them and put constrains on two of its atoms.<br>
I did a minimization and a simulation (md integrator).<br>It worked perfectly. But when I tried to perform the calculation<br>to free energy using the options below, the simulation<br>failed. Setting free-energy "no" the simulation works again.<br>
Do you have any suggestions? Bests <br>eef<br> <br>; Free energy control stuff<br>free-energy = yes<br>init-lambda = 1.0<br>delta-lambda = 0<br>sc-alpha = 0<br>sc-power = 0<br>
sc-sigma = 0.3<br><br>Error Message:<br>Back Off! I just backed up dgdl.xvg to ./#dgdl.xvg.9#<br>starting mdrun 'C60 - PMF'<br>30000 steps, 60.0 ps.<br><br>Back Off! I just backed up traj.trr to ./#traj.trr.18#<br>
step 0Warning: 1-4 interaction between 5 and 28 at distance 1.398 which is larger than the 1-4 table size 1.000 nm<br>These are ignored for the rest of the simulation<br>This usually means your system is exploding,<br>if not, you should increase table-extension in your mdp file<br>
<br><br><div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
<br>
Arthur Roberts wrote:<br>
> Dear Eudes,<br>
><br>
> You need to technically make it 1 "big molecule". As long as there are<br>
> no bonds specified between molecule A and B. The big molecule will<br>
> behave as if there are two molecules in it. I do it to dock small<br>
> molecules to proteins.<br>
<br>
That's a good description, Arthur, thanks.<br>
<br>
> 1) Combine molecules A and B into a big structure file (i.e. pdb or gro).<br>
> 2) Edit this file, so that they are the same chain.<br>
> 3) run pdb2gmx<br>
> 4) Change all the residue names to MOL and change Protein_A to MOL in<br>
> the top and pdb files.<br>
> 5) Make sure that there is no bonds between molecule A and B in the top<br>
> file. There probably won't be, but it is good to check anyway.<br>
> 5) run grompp to check this file to see if it is acceptable to gromacs.<br>
> 6) add distance restraints after the improper dihedrals. If you put it<br>
> anywhere else, it will not work.<br>
<br>
That contradicts the description of the topology file format on (e.g.)<br>
page 109 of the GROMACS 4 manual - "starting with GROMACS 3.1.3 all<br>
directives at the parameter level can be used multiple times and there<br>
are no restrictions on the order, except that an atom type must be<br>
defined before it can be used in other definitions."<br>
<br>
> 7) run grompp again to check this file to see if it is acceptable to gromacs<br>
><br>
> Other advise:<br>
> 1) I would roughly dock the molecules together, so that you don't have<br>
> to have such a large water box.<br>
> 2) I would experiment with force constants and time constants for the<br>
> distance restraints. A useful article is:<br>
<br>
Mark<br>
<br>
<br></blockquote></div><br>