[gmx-users] Doubts about g_lie and g_bar

[Mark Abraham]

Hi,

On Sat, May 13, 2017 at 5:18 PM Varvdekar Bhagyesh Rajendra <
bhagyesh.varvdekar at research.iiit.ac.in> wrote:

> Dear all,
>
> I have tried to find Delta G of Binding using g_lie and g_bar from the
> following systems (Protein+Ligand with ligand = chain B):
> PDB Ids: 1CDL, 1CKK, 1NIW, 2L7L, 2o60.
>
> I have encountered following 2 doubts while doing so:
>
> 1. This doubt may have been answered in the past threads but I couldn't
> find a precise answer for it. So, I would appreciate if this is answered
> here again in more details. I have used PME in the production runs of my
> systems for finding Delta G_Binding using g_lie. But many threads on
> Gromacs forum warn against it and advise to rerun the whole trajectory
> using Reaction field Zero, but the cutoff for this rerun is not clearly
> given. Some threads say a cutoff with large cutoff using RF-0 is necessary.
> Can anyone please answer as to what cutoff(Relative to the system) should
> be given while re-running the trajectory and what all changes have to be
> done in the original mdp file(which uses PME). Also, a brief explanation
> for the whole process for the need of the re-run would be beneficial.
> Following are the changes I made after studying the past threads:
>
> ------------------------------------------------------------------------------------------------------------------------
> ;; With PME  ;;;                                     | ;; Re-Run without
> PME ;;
>
> -------------------------------------------------------------------------------------------------------------------------
>  ; Neighborsearching                                 |  ; Neighborsearching
>   ns_type     = grid      ; search neighboring grid  |  ns_type     =
> grid      ; search neighboring grid cells
>   nstlist     = 5         ; 10 fs                    |  nstlist     = 5
>      ; 10 fs
>   rlist       = 0.9       ; short-range neighborlist |  rlist       = 1.2
>      ; short-range neighborlist cutoff (nm)
>   rcoulomb    = 0.9       ; short-range electrostatic|  rcoulomb    = 0.9
>      ; short-range electrostatic cutoff (in nm)
>   rvdw        = 1.4       ; short-range van der Waals|  rvdw        = 1.4
>      ; short-range van der Waals cutoff (in nm)
>                                                      |
> ; Electrostatics                                     | ; Electrostatics
> coulombtype     = PME                                | coulombtype  =
> Reaction-Field-zero
>                                                      | epsilon_rf      =  0
>
> ---------------------------------------------------------------------------------------------------------------------------
>

IMO these choices should follow from your understanding of the method that
you are implementing, e.g. as reported in peer-reviewed publications, not
just mailing list threads or tutorials. That way you have some confirmation
that the method was thought reasonable, and produced some kind of result.

2. Free Energy tutorials of Gromacs suggest using oriental restraints
> and/or distance restraints on protein-ligand systems for calculating Delta
> G_Binding using g_bar. But the exact details are not given on the website.
> One can see that in my systems the protein is fully surrounding the ligand
> except two entrances. So, I ask if distance and/or oriental constraint is
> necessary to keep from floating it away. If yes, can anyone suggest some
> simple ways for doing so in Gromacs. I found some research papers who
> employ gruesome techiques of finding the correct atoms to constraint in the
> ligand (for eg.Theoretical Study of the Binding Profile of the Allosteric
> Modulator NS-1738 with a Chimera Structure of the α 7 Nicotinic
> Acetylcholine Receptor).
>

The [intermolecular_interactions] directive introduced in recent GROMACS
versions is useful for such restraints, ie. when you want to refer to atoms
in different [moleculetypes], and can do so with global atom numbers.

Mark

I would be grateful if you could answer the above doubts.
>
> Thanking you in anticipation,
>
> Best Regards,
>
> Bhagyesh Varvdekar,
> Undergrad student,
> IIIT Hyderabad, India.
> --
> 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-request at gromacs.org.