[gmx-users] gmx wham problem

[edesantis]

Dear Matthew,
thanks for your opinion.
How can you establish if the histograms are sufficiently overlapped? Is 
there any thumbs role?
For what concern the negative sign of reaction coordinates, I think it 
could derive from the choice of the order of the pulling groups and the 
vector along you are pulling.

Thank you again
Best regards,
Emiliano


On 2017-06-28 17:49, Thompson, Matthew White wrote:
> It looks like you need to sample more states, 13 is not enough.
> Probably more like 20-30+ would be needed to get a smooth PMF as is
> discussed in that tutorial. The weird features in your PMF are from
> insufficiently overlapping histograms, for example the bump near -2.2
> nm corresponds to having no histogram there. You also see that you
> only have one state near -1.2 nm, so that is probably not being
> sampled enough for WHAM to produce meaningful results.
> 
> Also I don't understand the meaning of negative distance as a reaction
> coordinate. If that is the distance between two things, maybe it
> should be positive. It makes it difficult to understand which values
> correspond to them being close and far away.
> ________________________________________
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se
> [gromacs.org_gmx-users-bounces at maillist.sys.kth.se] on behalf of
> edesantis [edesantis at roma2.infn.it]
> Sent: Wednesday, June 28, 2017 10:26 AM
> To: Gmx users
> Subject: [gmx-users] gmx wham problem
> 
> dear all,
> 
> I am studying the affinity between an antibody and an amyloid peptide; 
> I
> am interested in the evaluation of the PMF. I have a problem with the
> PFM shape.
> I followed the protocol described in the umbrella-sampling tutorial
> (http://bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/umbrella/01_pdb2gmx.html)
> Here below there is the .mdp part for the pulling
> ; Pull code
> pull                    = yes
> pull_ngroups            = 2
> pull_ncoords            = 1
> pull_group1_name        = Chain_Abeta
> pull_group2_name        = Chains_Antibody
> pull_coord1_type        = umbrella      ; harmonic biasing force
> pull_coord1_geometry    = direction
> pull_coord1_groups      = 1 2
> pull_coord1_vec         =  38.207 68.611 29.8493
> pull_coord1_rate        = 0.002        ; 0.01 nm per ps = 10 nm per ns
> pull_coord1_k           = 1000          ; kJ mol^-1 nm^-2
> pull_coord1_start       = yes           ; define initial COM distance >
> 0
> 
> After the pulling simulation, I’ve extracted 13 configuration; for each
> them, 36 ns of equilibration were performed. These are the mdp
> directives:
> ; Pull code
> pull                    = yes
> pull_ngroups            = 2
> pull_ncoords            = 1
> pull_group1_name        = Chain_Abeta
> pull_group2_name        = Chains_Antibody
> pull_coord1_type        = umbrella      ; harmonic biasing force
> pull_coord1_geometry    = direction
> pull_coord1_groups      = 1 2
> pull_coord1_vec         = 38.207 68.611 29.8493
> pull_coord1_rate        = 0.00
> pull_coord1_k           = 1000          ; kJ mol^-1 nm^-2
> pull_coord1_start       = yes           ; define initial COM distance >
> 0
> 
> Then I ran the wham command:
> Gmx wham –it list_tpr.dat –if list_pullf.dat -v  –b 20000 –o –hist
> And I’ve obtained the following pictures:
> http://i66.tinypic.com/11t5zdv.png
> http://i67.tinypic.com/30u8g8x.png
> Do you have any idea of why the pfm profile has this strange shape?
> Could it come from any kind of error I’ve made during the simulations?
> If there are not errors, it seems that the configurations in which the
> peptide is far from the antibody are more energetically favoured 
> respect
> to those in contact with the antibody, but I have some doubts about it…
> 
> Can you help me?
> Thank you in advance,
> best regards,
> Emiliano
> 
> 
> --
> Emiliano De Santis
> --
> 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.

-- 
Emiliano De Santis