[gmx-users] Doubt about Free Energy Calculations using g_bar

[Justin Lemkul]

On 6/12/17 11:47 AM, Varvdekar Bhagyesh Rajendra wrote:
> Dear Justin,
> 
> I have compiled the following free energy code for use in the g_bar program for finding Binding affinity of a Protein-ligand complex, with ligand being a peptide. I would appreciate if you could please verify it's correctness and perhaps suggest any valuable improvements, if deemed necessary.
> 
> I am not sure about the input in the parameters for: *sc-coul* and *couple-intramol* though.
> 
> Since couple-lambda0 = none and couple-lambda1 = vdw-q, I assume coulomb interpolation must be taking place, and sc-coul should be *yes*.
> 

Not necessarily.  Scaling Coulombic interactions linearly is what is more 
commonly done.

> Also, as my ligand is a peptide, I presume couple-intramol should also be *yes*.
> 

If you're dealing with a peptide, you shouldn't be doing an alchemical 
transformation.  It will probably never converge.  PMF or MM/PBSA are much 
better approaches for determining binding free energies for these types of systems.

-Justin

> 
> ; Free energy control
> 
> free_energy              = yes
> init_lambda_state        = 0
> delta_lambda             = 0
> calc_lambda_neighbors    = 1        ; only immediate neighboring windows
> ; Vectors of lambda specified here
> ; Each combination is an index that is retrieved from init_lambda_state for each simulation
> 
> ; init_lambda_state        0    1    2    3    4    5    6    7    8    9    10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40
> vdw_lambdas              = 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
> coul_lambdas             = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
> ; We are not transforming any bonded or restrained interactions
> bonded_lambdas           = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
> restraint_lambdas        = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
> mass_lambdas             = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
> temperature_lambdas      = 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
> 
> ; Options for the decoupling
> sc-alpha                 = 0.5
> sc-coul                  = ??       ; As linear interpolation of Coulomb takes place, I suppose it should be *yes*
> sc-power                 = 1.0
> sc-sigma                 = 0.3
> couple-moltype           = Protein_chain_B      ; name of moleculetype to decouple
> couple-lambda0           = none     ; interactions are turned off in the beginning
> couple-lambda1           = vdw-q    ; all interactions are on in the end
> couple-intramol          = ??       ; since my ligand is a peptide, I suppose the intramol interactions should be *yes*
> nstdhdl                  = 10
> 
> 
> Thanking in anticipation,
> 
> Best Regards,
> 
> Bhagyesh
> 
> 
> ----- Original Message -----
> From: "Justin Lemkul" <jalemkul at vt.edu>
> To: gmx-users at gromacs.org
> Sent: Sunday, June 11, 2017 7:01:26 AM
> Subject: Re: [gmx-users] Doubt about Free Energy Calculations using g_bar
> 
> On 6/10/17 6:15 AM, Varvdekar Bhagyesh Rajendra wrote:
>> Dear Justin,
>>
>> Just so we are on the same page, this means that if I don't touch the topology file and use the following mdp snippet, charges *are present* in the topology file for the ligand group and they are automatically set to zero while turning on vdW interactions (from 0 to 1) right ? Hence, there is no need to manually set charges to zero (The old style of doing the calculations), right?.
>>
> 
> Yes.  You can easily confirm this using energygrps between your transformed
> molecule and the rest of the system.  Coul-SR should be zero.
> 
>> Then the charges too, are turned off from 0 to 1? (1 state being the actual charges present in the topology file).
>>
> 
> Charges are turned *on* if they are defined as off in the A-state (lambda=0) and
> on in the B-state (lambda=1).
> 
> -Justin
> 

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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