Re: [AMBER] Discrepancy between interaction energy of amber coordinate trajectories of a system

From: Carlos Simmerling <carlos.simmerling.gmail.com>
Date: Mon, 9 Nov 2020 09:51:55 -0500

I cannot comment on what VMD may have changed in the coordinates, but with
limited precision in ASCII crd files it would be expected that things like
the image and center commands that move atoms would produce slightly
different values for things like interatomic distances. Using binary netcdf
files is likely to reduce this problem.

On Sun, Nov 8, 2020 at 10:41 PM Prathit Chatterjee <
prathit.biophysics.gmail.com> wrote:

> Dear Prof. Simmerling and Dr. Ross,
>
> Sorry to reply to you back from the embedded amber digest, I did not
> receive separate reply emails.
>
> The precision of numbers is the same in both cases, and the format is
> .crd, please see below:
>
> *This is for the original amber coordinates*:
> parm prot.top
> trajin /homes/epsilon/users/diem/project_t-ab/traj_1/production/crd/
> *T_310K_1_md001.crd* 1 10
>
> #centering
> strip :WAT
> strip :Na+
> strip :Cl-
>
> center :1-43 mass origin
> image origin center familiar
>
> pairwise IE :1-43 out pairwise_IE.out cuteelec 0.0001 cutevdw 0.0001
> vmapout IE_vmap.out emapout IE_emap.out avgout IE_avg.out
> --------------
>
> *This is for the protein only generated coordinates*:
> parm prot.top
> trajin /homes/epsilon/users/pchatterjee/vmd_anal_prep/diem_1/
> *T_310K_prot_diem_1_100ns.crd* 1 10
>
> pairwise IE :1-43 out pairwise_IE.out cuteelec 0.0001 cutevdw 0.0001
> vmapout IE_vmap.out emapout IE_emap.out avgout IE_avg.out
> -------------
>
> Also, *I generated the protein-only amber coordinates as follows in VMD*:
>
> mol new
> /homes/epsilon/users/diem/project_t-ab/traj_10/production/whole.top type
> parm7
> mol addfile
> /homes/epsilon/users/diem/project_t-ab/traj_10/production/crd/T_310K_10_md$n.crd
> *type netcdf* waitfor all
>
> set frames [molinfo top get numframes]
> set nf [expr {[molinfo top get numframes]-1}]
>
> set a [atomselect top protein]
> animate *write crdbox* diem_10/PHF43_310K_prot_$f.crd beg 0 end
> $nf waitfor all sel $a
>
> I tried to write the protein only coordinates in several ways until
> reaching the above mentioned procedure, loading the original coordinates as
> netcdf, and writing the final protein only coordinates as crdbox.
> Likewise, the coordinates for visualizing the original and protein-only
> coordinates could be visualized properly without any apparent unnatural
> bonds.
>
> Best Regards,
> Prathit
>
>
> On Sun, Nov 8, 2020 at 5:00 AM <amber-request.ambermd.org> wrote:
>
>>
>> Message: 3
>> Date: Sat, 7 Nov 2020 05:28:00 -0800
>> From: Bill Ross <ross.cgl.ucsf.edu>
>> Subject: Re: [AMBER] Discrepancy between interaction energy of amber
>> coordinate trajectories of a system
>> To: amber.ambermd.org
>> Message-ID: <d455da94-5267-c71a-ed48-56c1e16ef604.cgl.ucsf.edu>
>> Content-Type: text/plain; charset=utf-8; format=flowed
>>
>> I wonder if the precision of the numbers is the same in both cases? What
>> format is your mdcrd in? If ASCII, precision is much reduced, otherwise
>> less of an issue but worth checking by e.g. calcing yourself with the
>> operations in a different order.
>>
>> Bill
>>
>> ------------------------------
>>
>> Message: 4
>> Date: Sat, 7 Nov 2020 09:44:44 -0500
>> From: Carlos Simmerling <carlos.simmerling.gmail.com>
>> Subject: Re: [AMBER] Discrepancy between interaction energy of amber
>> coordinate trajectories of a system
>> To: AMBER Mailing List <amber.ambermd.org>
>> Message-ID:
>> <CAGk3s-QY5CtFHALO95DVCN6kzcMmi-ePv4SjzxN=
>> 9pXo3vX1Mw.mail.gmail.com>
>> Content-Type: text/plain; charset="UTF-8"
>>
>> You haven't said much about how you actually did these steps, but I can
>> imagine that making a new (protein trajectory in ascii format with limited
>> precision could lead to small changes in coordinates and energy
>> differences.
>>
>> On Sat, Nov 7, 2020, 8:19 AM Prathit Chatterjee <
>> prathit.biophysics.gmail.com> wrote:
>>
>> > Dear AMBER experts,
>> >
>> > I had compiled a protein only amber trajectory of an explicitly solvated
>> > simulation with VMD, which I was able to visualize properly and run
>> certain
>> > analyses with it in VMD as well.
>> >
>> > But, for cross-checking, I tried to calculate the pairwise interaction
>> of
>> > the protein only generated trajectory (with VMD), and the original AMBER
>> > coordinates.
>> >
>> > The results, although very similar, are showing certain discrepancy as
>> > follows:
>> >
>> > [pchatterjee.node51 pairwise_IE]$ tail -f test*/pairwise_IE.out
>> > *==> test1/pairwise_IE.out <== [THIS IS THE ORIGINAL AMBER COORDINATE]*
>> > 1 -186.3642 -2973.0974
>> > 2 -189.9812 -2966.1259
>> > 3 -199.5370 -2957.7041
>> > 4 -196.3846 -2983.7931
>> > 5 -197.8264 -2981.1530
>> > 6 -184.0711 -2971.6430
>> > 7 -194.3659 -2970.1477
>> > 8 -189.0441 -2956.9308
>> > 9 -197.0864 -2948.2561
>> > 10 -196.4939 -3010.1367
>> >
>> > *==> test2/pairwise_IE.out <==[THIS IS THE PROTEIN ONLY GENERATED AMBER
>> > COORIDNATE IN VMD]*
>> > 1 -186.3943 -2972.9815
>> > 2 -189.9798 -2966.1573
>> > 3 -199.5455 -2957.6502
>> > 4 -196.3961 -2983.7881
>> > 5 -197.8009 -2981.1609
>> > 6 -184.0410 -2971.7176
>> > 7 -194.3681 -2970.0830
>> > 8 -189.0185 -2956.9181
>> > 9 -197.1040 -2948.2302
>> > 10 -196.4685 -3010.0947
>> >
>> > Although the global picture will be the same with such minor
>> discrepancy, I
>> > just want to know for the sake of knowing what the underlying reason
>> could
>> > be.
>> >
>> > Any comments/suggestions will be extremely helpful.
>> >
>> > Thank you in advance,
>> > Best Regards,
>> > Prathit
>> > _______________________________________________
>> > 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 Mon Nov 09 2020 - 07:00:02 PST
Custom Search