[Chimera-users] Fwd: Re: Bug (?) with EnsembleCluster

Francesco Pietra chiendarret at yahoo.com
Wed Jan 16 10:35:52 PST 2008

--- Elaine Meng <meng at cgl.ucsf.edu> wrote:

> Hi Francesco,
> There is no feature yet in Chimera to plot values (such as RMSD) vs.  
> time for a trajectory, sorry.

I hope you may plan to offer such useful tool.

> If you look at just the top tiny line in the RMSD map (frame 1 vs.  
> all the others) it contains that information in terms of color,

I may have better waited until I understand what you suggest. It is the hope to
further clarify my things that urges me to ask for that explained with other
words. I don't understand where to look at.

> but I  
> agree that is not as easy to look at as a plot.
> I don't know what happened in VMD; maybe there was some error.
> It is difficult to say what value of RMSD would correspond to too  
> much variation.  The values you gave don't sound that high to me, but  
> probably your judgment when viewing the trajectory is a better tool  
> than some numerical cutoff for deciding if it is reasonable or not.

That is reassuring. As I said, things "at the naked eye" look like OK. And the
judgment is competent because the ligand is an organic structure I am familiar
with. No repeating unit, special bond lengths and orientations, no common

> Does the interaction observed in your trajectory agree with any  
> experimental data on what residues contact the ligand?  Or if you  
> started with an experimentally determined structure of the complex,  
> does it stay in approximately the same binding site? (sounds like it  
> does)  Another thing to consider is that often people start with a  
> minimization to remove any severe strains or clashes, then some  
> equilibration MD before the production MD.  It depends on whether  
> your system is already equilibrated,

It was carefully heated to 300K and equilibrated removing restraints step by
step, i.e., first by the lipid, then by the protein-complex. If anything, I am
not sure if my procedure of removing the strong restraint on the protein and
ligand abruptly is correct, or if a smaller force should have been applied
first. It is not easy to grasp such information from the literature.

> or perhaps if the RMSD just  
> keeps increasing throughout the trajectory,

I am acquiring experience on these affairs. As far as I presently understand,
RMSD is not increasing along the trajectory. Well, a plot of RMSD vs time would
be the best answer.


> the system is unstable  
> and will not equilibrate.
> The following I mainly mention for completeness... it may not be  
> helpful in your case:
> There is another tool, EnsembleMatch, that shows the RMSD values as  
> numbers (instead of a color map) for an all-by-all comparison of two  
> ensembles, but it is only good for much smaller sets of structures.   
> For example, you could compare one structure (frame 1, or crystal  
> structure if you have it) to a sample of maybe 10 structures from the  
> trajectory.
> http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/ 
> ensemblematch/ensemblematch.html
> Best,
> Elaine
> -----
> Elaine C. Meng, Ph.D.                          meng at cgl.ucsf.edu
> UCSF Computer Graphics Lab and Babbitt Lab
> Department of Pharmaceutical Chemistry
> University of California, San Francisco
>                       http://www.cgl.ucsf.edu/home/meng/index.html
> On Jan 16, 2008, at 8:19 AM, Francesco Pietra wrote:
> > Yes, the two behave differently, and not only as to importing pdb.
> >
> > I had opened my combined trajectories (protein+ligand in a lipid  
> > membrane) in
> > VMD while looking at either the ligand or the protein: both highly  
> > distorted at
> > each one of the 550 frames. I only got nearly reasonable structure  
> > for the
> > ligand by averaging the frames. That was most discouraging in the  
> > last few
> > days. Did the same for the equilibration and got the same  
> > discouraging results.
> > This last analysis was not fitting the analysis with ptraj,
> >
> > Therefore I have now done the same for the MD trajectories with  
> > Chimera, which
> > runs more slowly along the frames, and surprisingly, "at the naked  
> > eye" both
> > the ligand and the protein save their correct structure along the  
> > 550 frames,
> > with only the internal displacement on/back that one expects from  
> > MD. I
> > repeated without hiding the membrane and water, and "at the naked"  
> > the ligand
> > was not wandering around.
> >
> > I looked at RMDS Maps(Start frame 1; End 550; Step size 2; RMSD map of
> > trajectory against itself; Lower rmsd threshold 2.9, higher 3.9:  
> > RMSD varied
> > from 0.194 to 1.810 for the ligand and from 0.612 to 2.420 for the  
> > protein. Is
> > that variation too much for accepting this MD for good (or fair)  
> > and continuing
> > it with same settings? If it is a too large a range, that explains  
> > why the
> > ligand encountered so many protein residues when the mask was used.
> >
> > I was unable to find how to get a plot of RMSD against time. I  
> > would like to do
> > that for the ligand (which is a single big residue) and selected  
> > residues of
> > the protein.
> >
> > Thanks
> > francesco

Never miss a thing.  Make Yahoo your home page. 

More information about the Chimera-users mailing list