Hello All,
I am back with my concerns about my free energy calculation project using TI (AMBER20).
So, in last 3 days I tweaked some parameters and I got different results and I am not sure which one to believe or rather how many calculations do I need to perform to assess the confidence of my results:
System ABL Kinase + isopropanol (ligand)
I changed solvent from TIP3P to SPCE and my free energy of complex changed from -2.80 kcal/mol to -3.60 kcal/mol
I ran the the same set of TI runs 10 different times and my net free energies change in the scale from +0.8 to -1 kcal/mol..
How to estimate error bars?
Thanks
Debarati
From: Debarati DasGupta<mailto:debarati_dasgupta.hotmail.com>
Sent: 07 October 2020 19:03
To: AMBER Mailing List<mailto:amber.ambermd.org>; Daniel Roe<mailto:daniel.r.roe.gmail.com>
Subject: Re: [AMBER] RMSD of ligand in pocket
Hello Dan,
Thanks for the reply.
Yes I did use ":2,23,271,274" as my restraintmask and my out file says that is matching 72 atoms. SO I cross checked it.
Also will try to try out both your methods and see which one makes more sense and can be doable.
But thanks again for the valuable suggestions!
Regards
Debarati
From: Daniel Roe<mailto:daniel.r.roe.gmail.com>
Sent: 07 October 2020 11:02
To: AMBER Mailing List<mailto:amber.ambermd.org>
Subject: Re: [AMBER] RMSD of ligand in pocket
Hi,
On Mon, Oct 5, 2020 at 8:17 AM Debarati DasGupta
<debarati_dasgupta.hotmail.com> wrote:
>
> Hello Dr. Case,
> Yes your right, we had discussed it a while back and so this time I reran my simulations (12 lambda 12 runs) with a restraint on my acetonitrile+ 5 protein residues surrounding it.
> (ntr =1, restraintmask=’2,23,271,274’) 274 is my C3N and the others are protein residues...
That's not a valid restraint mask - did you instead use something like
":2,23,271,274"? Even if so, I suspect that is far too few residues to
really restrain what you want (again, visualize the system to see what
the ligand is *actually* doing with respect to the receptor).
Two things you can try:
1) (Not optimal IMO) Fix your restraint mask so you're restraining
almost the entire protein backbone and the ligand, something like
"(:2-271.C,CA,N)|(:274&!.H=)" (backbone of residues 2 to 271, residue
274 and not hydrogens).
2) (Better IMO) Use 1 or more distance restraints (see the "Distance,
angle and torsional restraints" section of the manual) to restrain the
movement of the ligand w.r.t. the protein. This will allow the protein
and ligand to move more while still maintaining close contact (and
ergo probably introducing less bias into your system).
Keep in mind in either case that when you are analyzing the data you
need to take into account the fact that you're introducing external
restraints and therefore modifying the system behavior. So you may
need to e.g. reweight the data.
-Dan
> So, even if the restraint is applied to not just the ligand but the ligand and the residues around it, even then how does it meander away to like 4-5 Angstroms from the initial site location?
> Am I doing it wrong ?
> May you kindly suggest me some ideas on how to make sure the ligand does not walk away from the protein or vice versa.
> Thanks so much sir.
> Debarati
>
>
>
>
>
> Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10
>
> From: David A Case<mailto:david.case.rutgers.edu>
> Sent: 04 October 2020 20:54
> To: AMBER Mailing List<mailto:amber.ambermd.org>
> Subject: Re: [AMBER] RMSD of ligand in pocket
>
> On Sun, Oct 04, 2020, Debarati DasGupta wrote:
> >
> >While applying a restraint of 10kcal/mol on my C3N ligand, why does the
> >rms fluctuations look like this...( Plot of 12 lambdas C3N RMSD over
> >the production phase of TI runs)???
>
> Is this perhaps the problem that was discussed a while ago? Are you
> restraining the *ligand* to stay in its original location, but allowing
> the protein to move? In that case, the restraint on the ligand
> essentially has no effect at all: letting the protein move away from the
> ligand is equivalent to letting the ligand move away from the protein.
>
> If you then run your cpptraj script, which superimposes the protein
> frames onto a reference, the ligand will appear to be moving all over
> the place.
>
> Visualization of one of the trajctories is a must here (if you've not
> already done that.) You can see what the ligand is doing. If it's
> moving out of the pocket, there's nothing that cpptraj commands can do
> after the fact.
>
> If you want restraints to keep the ligand in the pocket, then need to be
> relative restraints connecting the ligand and the protein, not
> "absolute" ones that just keep the ligand (only) in one place.
>
> Apologies if I'm off base here or mis-remembering things!
>
> ....dac
>
>
> _______________________________________________
> 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
_______________________________________________
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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Oct 18 2020 - 08:30:02 PDT
