I'm not sure if I understand exactly what you did, but note that rigid
water still has 6 degrees of freedom and only 3 have been removed. it
wasn't clear if you *subtracted* 6 for each water (this is what I think you
meant) or *kept* 6 per water.
On Wed, Jul 17, 2024 at 7:46 AM Zihan Zhou <zihan.zhou.uq.edu.au> wrote:
> Hi Carlos,
>
> Thanks for the reply.
>
> Yes I understand the temperature in the mdin is the target temperature,
> which I set to 333 K.
>
> The temperature in the mdout is consistent with those in mden (output
> energy file, in the AMBER2022 documentation this is described as "mden
> output extensive energy data over trajectory" and was written with the -e
> flag when calling pmemd or sander), this is what I was referring to when I
> wrote "a. temperatures written in the .mden file" and this showing a system
> temperature of ~333 K averaged across frames for which .mden output was
> written.
>
> And yes I did take the number of constrained degrees of freedom I have in
> the system into account when calculating the temperature from kinetic
> energy and velocity. When using the "temperature" action I did this by
> specifying the level of constraints using the "ntc" option, I used ntc=2
> (bonds containing H atoms constrained) which is consistent with what I set
> in the control file (mdin). When calculating the temperature from the
> kinetic energy myself, the degree of freedom I used was 3N-number of
> constrained DOF - 3 with N being the number of atoms in the system, the
> number of constrained degrees of freedom is the number of H containing
> bonds in the solute molecules + 6*number of solvent molecules. The solvent
> molecule used was rigid water so it should have 6 degrees of freedom for
> each molecule.
>
> Hope this clears up my question.
>
> You mentioned that the temperature reported in the mdout is the most
> direct way to get the temperature, I still am unclear on how these values
> are obtained and why can I not reproduce that temperature value using
> velocities or kinetic energies like in the other two ways I tried (that is,
> in b. and c.)?
>
> Regards,
> Zihan
>
> ------------------------------
> *From:* Carlos Simmerling <carlos.simmerling.gmail.com>
> *Sent:* 17 July 2024 20:50
> *To:* Zihan Zhou <zihan.zhou.uq.edu.au>; AMBER Mailing List <
> amber.ambermd.org>
> *Subject:* Re: [AMBER] Temperature of a system
>
> The temperature in mdin is the target T of the thermostat.
> When you use the KE or velocities, you need to ask use the correct number
> of degrees of freedom. This must account for use of shake. You didn't
> mention if you did this or not.
> Also I see you are using the Berendsen thermostat with quite weak
> coupling. This model is less commonly used these days.
> Also you don't mention the temperature reported in the mdout, which is
> the most direct way to get what you want.
>
> On Wed, Jul 17, 2024, 2:21 AM Zihan Zhou via AMBER <amber.ambermd.org>
> wrote:
>
> Hi all,
>
> I have been trying to find out the temperature of a system I ran with
> AMBER2022 (PMEMD implementation of SANDER, Release 22).
>
> To do so, I obtained the temperature from 3 different ways:
>
> a. temperatures written in the .mden file
> b. calculating temperature from the velocities written (using the
> "temperature" action of cpptraj)
> c. calculating temperature from kinetic energy myself, with the kinetic
> energy obtained using "energy" action of cpptraj (specifically using the
> ketype vv option, which apparently results in cpptraj "Assuming plus-half
> time step velocities. Velocities at time 't' will be estimated using force
> info")
>
> I am a bit perplexed to find that the temperature obtained as described
> above all differ from each other.
> with a. giving an average temperature of ~333 K, b. giving an average
> temperature of ~342 K and c. giving an average temperature of ~ 344 K
>
> So I am wondering how should I obtain temperature of my system?
> Also how is the temperature written in the .mden file actually determined
> and why does it differ from the temperature obtained with the other ways so
> significantly ?
>
> ===below is the run control (mdin) details of my simulation===
> barostat=1,
> imin=0, ntx=1, irest=0,
> ntpr=500, ntwr=50000, ntwx=500, ntwv=-1, ntwe=500, ntwf=-1
> ntf=2,ntb=2, nsnb=5,cut=10.0,
> nstlim=500000, nscm=100, t = 0.00000, dt = 0.002, vlimit
> = 0.
> ntt=1, ig=-1, temp0 = 333.00000, tempi = 333, tautp= 0.1,
> ntp = 3, pres0 = 1.00000, comp = 46.00000, taup =
> 1.0000
> csurften = 3, gamma_ten = 0.00000, ninterface = 2,
> ntc=2, jfastw=0, tol=0.0004,
> ene_avg_sampling=100,
> &end
> &ewald
> ew_type= 0,
> nbflag=0,
> vdwmeth=0,
> eedmeth=1,
> netfrc=1,
> &end
> ===================================================
>
> Regards,
> Zihan
> _______________________________________________
> 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 Jul 17 2024 - 11:30:04 PDT
