----- Original Message ----- From: Javier Cerezo <j...@um.es> Date: Thursday, September 9, 2010 3:17 Subject: [gmx-users] dihedrals potential as a expansion of cosine funtions To: gmx-users@gromacs.org
> Hi all. > > I am a bit confused about how GROMACS handles multiply defined > interactions in the top file. Concretely, I see from the lipid > parameters (POPC) from Kukol (JCTC, 5:615, updated in > http://lists.gromacs.org/pipermail/gmx-users/2010- > May/050647.html ) that when defining several times an > interaction it is sum up. However, it was also stated by Mark > Abraham that just the last definition in the top file is taken > into account (http://lists.gromacs.org/pipermail/gmx-users/2007- > January/025316.html):Quote: > > << > For purposes of GROMACS, a bonded function is defined by the > function type > (an integer) and the set of atoms to which it applies. See table 5.4. > There may only be one instance of a given function type for a > given set of > atoms. grompp will take the last such function that it finds. > Thus you > cannot have multiple periodic dihedral functions on the same > four atoms, > even if their multiplicity varies. > >> > > > Could anyone help me with that? How will GROMACS-4 interpret > multiply defined interaction? GROMACS behaves differently when looking up parameters for a dihedral function from the [dihedraltypes] than when assembling dihedral functions whose parameters are already supplied in [dihedrals]. (Some time since 2007 you will find a post from Berk that corrects me on this point :-)) The GROMOS forcefields use the latter method via the gb_xx cpp macros. These add. You can verify this simply by adding a duplicate of a single [dihedral] and observing that the Step 0 energy will change. Normal [dihedraltypes] of the same "kind" do not add, except for type 9 (introduced in GROMACS 4.5) which caters to aspects of CHARMM and AMBER force fields. Now that I think about it, perhaps a sentence is needed in the manual treating this case. Will check there. Mark
-- gmx-users mailing list gmx-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
