How to check extent of convergence, and how to use developed potentials in a new run?

[Satyen Dhamankar]

Hello,

I am running IBI on a particular polymer simulation, and I VOTCA has some files in step_xxx directories. My question is, given a distribution I have asked to recreate, A-B.dist.tgt, how do I check the extent to which it has been reproduced in the simulation? 

For example, I am trying to optimize TW-TW.dist.tgt, and these are all the TW-TW files that have been generated: 

TW-TW.conv
TW-TW.dist.conv
TW-TW.dist.new
TW-TW.dist.new.resample.fnF
TW-TW.dist.tgt
TW-TW.dist.tgt_extrapolated.hpVpl
TW-TW.dist.tgt_extrapolated_left.fLt6f
TW-TW.dist.tgt_extrapolated_left.RchsI
TW-TW.dist.tgt_extrapolated.ly2nG
TW-TW.dist.tgt_smooth.TEwdo
TW-TW.dist.tgt_smooth.TiJo0
TW-TW.dpot.1
TW-TW.dpot.new
TW-TW.dpot.pure_ibi
TW-TW.pot.1
TW-TW.pot.2
TW-TW.pot.3
TW-TW.pot.cur
TW-TW.pot.extrapol.8Lku6
TW-TW.pot.interpol.P4rlN
TW-TW.pot.new
TW-TW.pot.shift.QoRc9
TW-TW.pot.smooth.hnK0t
TW-TW.pot.smooth.hnK0t.onlyleft.zhxPa

Furthermore, in the event that IBI has converged successfully, are the tables, table_TW_TW.xvg what I am supposed to use for a long-scale simulation?

 

Jumping off of the question above, if I want to run an IBI where I want to recreate other potentials and use the potentials developed by a previous IBI. For example, running an IBI on bonded potentials after running an IBI on non-bonded potentials, how do I do this on VOTCA? 

[Christoph Junghans]

On Wed, Oct 6, 2021 at 9:04 AM Satyen Dhamankar <saty...@gmail.com> wrote:
>
> Hello,
>
> I am running IBI on a particular polymer simulation, and I VOTCA has some files in step_xxx directories. My question is, given a distribution I have asked to recreate, A-B.dist.tgt, how do I check the extent to which it has been reproduced in the simulation?

dist.new is the file to check, it is the distribution done with the
current potential. (.pot.cur)

Correct!

>
> Jumping off of the question above, if I want to run an IBI where I want to recreate other potentials and use the potentials developed by a previous IBI. For example, running an IBI on bonded potentials after running an IBI on non-bonded potentials, how do I do this on VOTCA?

If you want to iterate on them more, you can put a <name>.pot.in in
the main directory as an initial guess.
If you don't want to iterate on them more, you just add the final xvg
file to <filelist> and remove their block from settings.xml.

Christoph
>
> --
> Join us on Slack: https://join.slack.com/t/votca/signup
> ---
> You received this message because you are subscribed to the Google Groups "votca" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to votca+un...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/votca/8c4cfe4b-9ab3-44d3-8cc7-9c3afe031720n%40googlegroups.com.



--
Christoph Junghans
Web: http://www.compphys.de

[Satyen Dhamankar]

Thank you for your response Christoph! 

I also have a quick question about performance. 

These simulations consisting of about 11000 atoms (not a very large system), is taking much longer than my all-atom simulation. I am seeing the following message in when mdrun is run on GMX:

"Multiple energy groups is not implemented for GPUs, falling back to the CPU. For better performance, run on the GPU without energy groups and then do gmx mdrun -rerun option of the trajectory with an energy group .tpr file." 

Is the lack of GPU support leading to about 0.1 ns/hr speeds? Are tabulated potentials for non-bonded/bonded interactions slower in GMX? 

[Christoph Junghans]

On Thu, Oct 7, 2021 at 9:48 AM Satyen Dhamankar <saty...@gmail.com> wrote:
>
> Thank you for your response Christoph!
>
> I also have a quick question about performance.
>
> These simulations consisting of about 11000 atoms (not a very large system), is taking much longer than my all-atom simulation. I am seeing the following message in when mdrun is run on GMX:
>
> "Multiple energy groups is not implemented for GPUs, falling back to the CPU. For better performance, run on the GPU without energy groups and then do gmx mdrun -rerun option of the trajectory with an energy group .tpr file."
>
> Is the lack of GPU support leading to about 0.1 ns/hr speeds? Are tabulated potentials for non-bonded/bonded interactions slower in GMX?

If you have a GPU in your machine that is correct. There is not much
we on the VOTCA side can do here, but wait.
Please free to ping the gromacs developers for a status update here:
https://gitlab.com/gromacs/gromacs/-/issues/1347

Christoph

> To view this discussion on the web visit https://groups.google.com/d/msgid/votca/a5d36399-0df3-40f8-9741-7f12e958fdc2n%40googlegroups.com.