Dear Amber Community:
I am studying a protein with an exposed surface loop. I am interested in
whether the loop could deform and interact with another part of the
protein. As a "quick and dirty" method to investigate the possibility, I
have tried to do an energy minimization in a GB simulation with a distance
restraint involving an increasing force constant as a function of
minimization step number to bring the loop and potential interaction site
close together. The approach is not working. I have increased the force
constant by orders of magnitude and the loop never appears to be moving
toward the potential interaction site.
I have several questions.
1. Is this approach even feasible in Amber8? What am I missing?
2. If it is, could someone help me understand what I am doing wrong?
I am using Amber8. The min.in file follows:
minimize structure
&cntrl
imin=1,
ntmin=1,
ncyc=200,
maxcyc=200,
cut=300.0,
igb=2,
saltcon=0.2,
gbsa=1,
ntpr=10,
ntx=1,
ntb=0,
ntr=1,
nmropt=1,
lastist=30000000,
lastrst=30000000,
&end
&wt
type='END',
&end
LISTOUT=/home/pate/david_eg5.d/listout
DISANG=/home/pate/david_eg5.d/RST
restrain the following
5.0
RES 1 117
RES 133 500
END
END
The loop in question is residues 118-132. The distance restraint on the
rest of the protein was an effort to try and keep the protein from being
pulled apart during the restraint minimization. Elimination of this group
definition does not affect the end result.
The RST file is
&rst
ixpk=0, nxpk =0, iat(1)=1971, iat(2)=5834, iat(3)=0, iat(4)=1,
r1=1.5, r2=2.5, r3=3.5, r4=45., r1a=1.5, r2a=2.5, r3a=3.5, r4a=45.,
rk2=0.1, rk3=0.1, rk2a=10000000., rk3a=10000000.,
ir6=0, ialtd=0, ifvari=1, ninc=0, iresid=0, imult=0,
nstep1=0, nstep2=201,
&end
Atom 1971 is in the loop. Atom 5834 is in a ligand bound to the protein.
It is in residue 387. The .top and .crd files for the protein were
generated using the outline in the tutorial at
http://ambermd.org/tutorial/gb/gb_tutorial.html. tleap reported no
problems. Atoms/residues were identified using ptraj to generate a pdb
file. Atoms 1971 and 5834 were initially approx. 25A apart. After the
iteration, they remained approx. the same distance apart. Smaller, more
reasonable values of rk2a and rk3a yield the same result.
If someone could help me understand whether the approach is feasible, and
if so, where I am going wrong, I would appreciate it.
Thanks,
Ed Pate
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jan 27 2010 - 21:30:02 PST
