Dear Xiaoyu,
Let me follow up on Dave's message and perhaps add a few things that
hopefully might be helpful.
Just quick to clarify - by "evaporation" I am assuming that you mean an
absolute solvation free energy calculation where you make your molecule
"disappear" (transform into a non-interacting dummy state) in the
solvent of interest and also in the gas phase, and take the difference
in those free energies - is this correct?
Here are a couple of quick comments I hope you find useful.
First - are you sure you want to use "traditional softcore potentials"
(which are not the default in AMBER 22 - the default, and what we
strongly recommend, is to use the latest smoothstep softcore potentials
described here: DOI: 10.1021/acs.jctc.2c00725). These softcore
potentials were designed such that you do not need to perform stepwise
decharge-vdW-recharge steps, but you can still do that if you want to.
To ensure these latest softcore potentials are used - make sure to set
the flags:
gti_cut = 1
gti_output = 1
scalpha = 0.5
scbeta = 1.0
gti_add_sc = 5
gti_scale_beta = 1
gti_cut_sc_on = 8
gti_cut_sc_off = 10
gti_ele_sc = 1
gti_vdw_sc = 1
gti_cut_sc = 2
gti_ele_exp = 2
gti_vdw_exp = 2
Second - regardless of what softcore potential you choose to use, if you
really want to do decharge-vdW-recharge, there are two ways that you can
do it. The first is to do it "by hand", as indicated by others in their
response, where you define uncharged intermediates as a separate
parameterized molecule and perform separate free energy simulations in a
stepwise fashion - as described here: DOI:10.1021/acs.jcim.0c00285. the
other way is to use the new "lambda-scheduling" that we implemented and
is described here (DOI: 10.1021/bk-2021-1397.ch007) and in our free
energy workflow paper (DOI: 10.1021/acs.jcim.2c00879). To use the lambda
scheduling procedure to perform decharge-vdW-recharge, you would need to
(1) set the control flag
gti_lam_sch = 1
(2) add one more option after the pmemd executable as -
pmemd ... -lambda_sch lambda.sch
(3) have a text file named lambda.sch with content as -
LambdaType, FunctionType, Matchtype, parameter1, parameter2
one example is -
TypeEleCC, smooth_step2, symmetric, 0.0, 1.0
More details can be found in Amber 2023 Manual section 25.1.7.1. If you
want us to try and put together some more detailed file examples, we can
try and do that for you.
Third, I might mention, depending on the complexity of your molecule, it
might be valuable to use the ACES enhanced sampling method during the
free energy perturbation - this helps to insure that the "dummy state"
that results from your alchemical transformations in the gas phase and
in solution actually sample the same free energy basin and can thus be
connected in the thermodynamic cycle. There is a tutorial (7.2) on how
to do this, and the paper tht describes the method is here: *DOI*:
10.1021/acs.jctc.2c00697.
Finally - there are a couple of reasons you may NOT want to use the
stepwise approach but rather use a concerted transformation where you
change the charge and vdW at the same time using the new properly
balanced smoothstep softcore potentials that avoid the traditional
problems of end-point catastrophe, particle collapse and large
gradient-jump. The first is that decharging can involve charge-changing
perturbations that sometimes lead to very large free energy changes - as
a simple motivating example, imagine Mg(II)->Ca(II). Second, this
procedure is not readily extendable to other contexts such as relative
binding free energy calculations, where often decharging a ligand will
result in it drifting out of the binding pocket if you do not impose
other restraints. The main motivation for doing decharge-vdW-recharge
in the first place is to loosen up the strong short-ranged electrostatic
interactions such that the free energy of perturbing the vdW
(Lennard-Jones terms) is more easily computed - but the new optimized
alchemical transformation pathways with smoothstep softcore potentials
achieve the same end in a different way that has many additional
benefits. Third, using a concerted approach makes it easier to
integrate enhanced sampling methods like ACES (that leverages the
Hamiltonian replica-exchange framework in Amber) into the free energy
simulation.
A recent perspective article that just came out in ACS Au Phys could be
useful to put some of these issues into better context: "Modern
Alchemical Free Energy Methods for Drug Discovery Explained",
https://pubs.acs.org/doi/10.1021/acsphyschemau.3c00033.
I hope this response was helpful!
best wishes,
Darrin York
On 10/4/2023 10:32 AM, David A Case via AMBER wrote:
> On Wed, Oct 04, 2023, Xiaoyu Wang via AMBER wrote:
>>
>> I am trying to set up a calculation on the evaporation free energy
>> estimation. As I went through manual, I saw that AMBER implements the
>> traditional soft-core potential for elec and vdw. But I did not find
>> the section which describes the instruction to decouple elec and vdw
>> separately. Can anyone guide me on this?
>
> As Thomas indicated, I don't think there is any "instruction" that does a
> decoupled vdW and electrostatic calculation. You probably have to run
> several jobs and combine the results.
>
> Maybe folks on the list that do this can provide an overview (or detailed
> files) of how they approach this.
>
> ....dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
======================================================================
Darrin M. York : Distinguished Professor
: Henry Rutgers University Chair
Institute for Quantitative :
Biomedicine and Department of : Director, Laboratory for
Chemistry & Chemical Biology : Biomolecular Simulation Research
:
Rutgers, the State University :Darrin.York.rutgers.edu
of New Jersey : fax: +1-732-445-4320
174 Frelinghuysen Road : phone: +1-848-445-5199
Piscataway, NJ 08854 USA :http://theory.rutgers.edu/
======================================================================
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Oct 04 2023 - 10:00:02 PDT
