Re: [AMBER] Thermodynamic Integration query

From: Sowmya Indrakumar <soemya.kemi.dtu.dk>
Date: Wed, 14 Feb 2018 09:19:53 +0000

Hi Daniel,

Thank you for your comments.

I'm interested in studying the free energy change in going from disulphide bond state to the unbound state.

Since I run simulation in pmend, I had to make a single parameter file.

To do this, I used timerge, it removes the atoms that are duplicated except the ones mentioned in the softcore mask (59, 79).

The masked residues are renumbered to (41, 42). Hence, I changed the numbering of these atoms during minimization.


Do you have an idea about the sensitivity of this method, i.e. can it measure dG difference of 1kCal or less ?


This is how my results look like:


0.0 29.247630214
0.1 21.9795085856 0.146961851401 14.6961851401
0.2 7.61081771445 0.140120093147 14.0120093147
0.3 2.57943509958 0.126734448996 12.6734448996
0.4 4.78184922302 0.116245345494 11.6245345494
0.5 6.01238819561 0.116237525479 11.6237525479
0.6 9.38179279008 0.127445262973 12.7445262973
0.7 3.40368163611 0.119908467485 11.9908467485
0.8 -0.581106464259 0.127984306045 12.7984306045
0.9 -18.8675387282 0.134582088125 13.4582088125
1.0 -81.4630559713
# dG = 1.01931151733


I'm not very sure what each of these columns corresponds too. I guess the second one is dv/dl, and the 3rd column is probably standard error of mean of the 4th column (standard deviation).

Do you think this calculation is good enough to say bonded state is preferred over the free cysteines?


Thanks in advance.

Regards

Sowmya


________________________________
From: Daniel Mermelstein <danmermelstein35.gmail.com>
Sent: Wednesday, February 14, 2018 7:03:47 AM
To: AMBER Mailing List
Subject: Re: [AMBER] Thermodynamic Integration query

Hi Sowmya,

This is an interesting simulation. Depending on what you're trying to get
out of the simulation, you may not need to add the HG atoms as dummy atoms.
If you define the cysteine residues you're changing as softcore, then it's
okay if they have a different number of atoms between region 1 and region
2.

On Fri, Feb 2, 2018 at 2:15 AM, Sowmya Indrakumar <soemya.kemi.dtu.dk>
wrote:

> Dear Amber users,
>
> I have run TI simulation on a system using pmemd (which needs single
> parmtop and inpcrd), I have in one state bonded disulphide bridge (state A)
> and in the other free cysteines (State B).
>
>
> As I understand, CYX is recognised as bonded and CYS as free cysteine in
> amber, hence the partial charges are adjusted accordingly. Since free
> cysteines have an extra atom (HG), should these be added as a dummy atom to
> the bonded cysteines??
>
>
> This is how I made the input files for running MD:
>
> 1) leap file
>
>
> source leaprc.protein.ff14SB
> source leaprc.water.tip3p
> loadAmberParams frcmod.ionsjc_tip3p
> m1 = loadpdb new_ss.pdb
> m2= loadpdb new.pdb
> bond m1.4.SG m1.30.SG
> bond m1.15.SG m1.37.SG
> bond m1.19.SG m1.39.SG
> bond m2.4.SG m2.30.SG
> bond m2.15.SG m2.37.SG
> protein = combine {m1 m2}
> set default nocenter on
> solvateoct protein TIP3PBOX 15.0
> savepdb protein protein_combine.pdb
> saveamberparm protein protein_sol.parm7 protein_sol.rst7
>
> quit
>
>
> 2) Timerge
>
> parmed protein_sol.parm7 <<_EOF
> loadRestrt protein_sol.rst7
> setOverwrite True
> tiMerge :1-40 :41-80 :19,39 :59,79
> outparm merged_protein.parm7 merged_protein.rst7
> quit
> _EOF
>
>
> 3)Minimisation using the merged_protein.parm7 merged_protein.rst7
>
> &cntrl
> imin = 1, ntmin = 2, maxcyc = 2000, ncyc=1000,
> ntpr = 20, ntwe = 20,
> dx0 = 1.0D-7,
> ntb = 1,
> iwrap = 1,
> icfe = 1, ifsc = 1, clambda = %L%, scalpha = 0.5, scbeta = 12.0,
> logdvdl = 0,
> timask1 = ':19,39',
> timask2 = ':41,42',
> scmask1 = ':19,39',
> scmask2 = ':41,42',
>

one possible issue here is you have residues 41 and 42 defined as softcore
for region 2, but in timerge you've specified scmask2 as residues 59 and
79. I haven't used timerge in a while so maybe I'm just not familiar, but
just make sure you've got the correct softcore masks.


>
> !method 2:
> !scmask1 = ':19,39', scmask2 = ':41,42',
> /
>
> &ewald
> /
>
> Looking forward to some comments on this.
> Thanks in advance
> Regards
> Sowmya
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>

Best,

Dan


--
Daniel J. Mermelstein, Ph.D candidate
McCammon & Walker Groups
Department of Chemistry & Biochemistry
University of California, San Diego
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 14 2018 - 01:30:02 PST
Custom Search