[gmx-users] problem in analyzing

[Justin A. Lemkul]

hazizian wrote:
> Hi
> I want to analyze md. This md took 10 ns(10000 ps) for a protein with 206 
> residue and 2 Mn plus about 17000 molecule solvent as water.
> then I construct extra group containig Protein+2 Mn with make_ndx comand.
> I did g_rmsf for extracting the average structure of the new group rmsd 
> equilibrated section :
> 
> g_rmsf -n p+mn.ndx -s md300.tpr -f md300.trr -b 9000 -e 10000 -ox e.pdb              
> then I did procheck to evaluate e.pdb, the summery of ramachandran was:
> 
> 83.3% core 15.2% allow 0.5% gener 1% disall  G-factor= -0.75
> 
> for improving this result I did energy minimization with steepest descent 
> and cg. the parameter for steep is emtol=1000 the ramachandran result was:
> 
> 82.8% core 15.7% allow 0% gener 1.5% disall  G-factor= -0.3
> 
> steep emtol=10:
> 76% core 22.1% allow 0.5% gener 1.5% disall  G-factor= -0.47
> 
> after performing cg nstcgstep= 9000 emtol=1 nstep=0.01 the result was
> 76% core 21.6% allow 1% gener 1.5% disall  G-factor= -0.48
> 
> it seems that the results get worse after doing EM.
> my question is why  performing EM worse this result? 

Realize that an "average structure" may have no physical significance at all:

http://wiki.gromacs.org/index.php/Average_Structure

> what could I to improve ramachandran result, is it better to do md for 
> longer time(according to 206 residue)?

Maybe, maybe not.  You might be better off analyzing your result with something 
like g_cluster to group structures by RMSD clustering to get a sense of the 
prevailing structures in the simulation.

> another question is whether incresing tempreture from 300 to 500 or greater 
> affect procheck result or not? besids changing temprature and nstep 
> parameter in md.mdp with chaning in which parameter can we improve md 
> results?

Depending on your protein's stability, it may just unfold, and that's probably 
not what you're after.

-Justin


> thank's for replying.
> 
> 
> 
> 
> 

-- 
========================================

Justin A. Lemkul
Graduate Research Assistant
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================