Re: [AMBER] Questions (and a small bug report) about SIRAH in Amber

From: Matias Machado <mmachado.pasteur.edu.uy>
Date: Mon, 24 Aug 2020 13:09:09 -0300 (UYT)

Dear Charo,

mmm... SIRAH failing on Linux that's something I've never seen...

I'm a Linux-only user as well, and all SIRAH packages were tested OK on different Linux DistrOS and HPC clusters...

Can you please tell me more about your OS distribution and the AMBER compilation? In particular, are you installing AMBER though conda?

The LJoff issue I know is relate to the file format and Leap... but the "buffer overflow" error is something new that I need to reproduce, otherwise I can't tell you what's going on there...

About your question: how do you know that the prmtop files obtained on a machine that does not report the bug are correct?

I never seen that to happen, not even in your case... one way or another Leap fails generating the topology... the fact that you managed to "overcame the error message" doesn't mean you haven't had an error in the first place...

Any way we will include some test cases during AMBER installation to check for these issues...

Jumping to the question of multiscaling...

Simply mixing FG and CG is highly expected to fail... there are several reasons for that, starting from incompatible interaction points, non-bonded parameters at very different energy references and so on... so far we provide some multiscale representations for DNA and solvent, but there's a long way before getting to proteins...

About the thermostats...

If by "Bussi thermostat" you are referring to V-rescale, then it should work, it's the one we use in GROMACS... Indeed I see no reason for any thermostat to fail...

All the best,

Matias

----- Mensaje original -----
De: "paraw" <the.paraw.gmail.com>
Para: "AMBER Mailing List" <amber.ambermd.org>
CC: "david case" <david.case.rutgers.edu>
Enviados: Lunes, 24 de Agosto 2020 12:19:42
Asunto: [AMBER] Questions (and a small bug report) about SIRAH in Amber

Dear Matias,
thank you very much for your reply.
As for the bug, I don't use MacOS. In fact, I only use Linux machines and
clusters (many of which I set up myself, in fact). Anyway, I'll wait for
more information about the workaround or a fix, but I wonder: how do you
know that the prmtop files obtained on a machine that does not report the
bug are correct? All you could say is that somehow the bug wasn't detected
or triggered there, and considering that we are talking about a buffer
overflow, the possibility of having overwritten some important memory
location cannot be excluded.

Finally, I didn't CG the ligand. In one simulation, I just kept it FG, and
in another I simply got rid of it and added a network of harmonic
restraints to keep the shape of the active site. Note that the simulation
without ligand was unstable with the Bussi thermostat but stable if I first
eqiilibrated it with Langevin dynamics and gamma_ln=50, and then passed to
Bussi. The one with the ligand, instead, was only stable if I first heated
it extremely slowly (dt=0.001) at constant volume with Langevin thermostat,
before passing to constant pressure.

So, my other questions remain: first, is SIRAH compatible with Bussi
thermostat, or does it need Langevin with high viscosity? Then, is the
strategy of leaving an atomistic ligand in a CG protein viable or not?


Thank you again,

Charo




----
Dr. Charo I. del Genio
Senior Lecturer in Statistical Physics
Applied Mathematics Research Centre (AMRC)
Design Hub
Coventry University Technology Park
Coventry CV1 5FB
UK
https://charodelgenio.weebly.com
On Mon, 24 Aug 2020, 15:27 Matias Machado, <mmachado.pasteur.edu.uy> wrote:
>
> Dear Charo,
>
> Sorry for the delay answer, I'm recovering from a high fiver (luckily not
COVID-19)... :-P
>
> Any way, going to your questions:
>
> 1) As you point, SIRAH is not in ${AMBERHOME}/dat/SIRAH, I've just
written to Sergio Pantano (the group leader) about the issue because I
don't have the GitLab credentials to fix this at the AMBER's repo...
>
> 2) SIRAH is a stand alone package, so once downloaded it should work at
any location as long as you correctly set the "addPath" command (check any
AMBER tutorial at http://www.siraff.com)
>
> 3) THE FOLLOWING IS EXTREMELY IMPORTANT!
> We are aware of the bug you mention about "LJoff.frcmod"... so far we
have only seen the issue in some MacOS, which may be your case... We are
digging to fix it, I have some leads pointing either to the OS version
(probably Catalina) or to the environment (perhaps Python)... but it's
being hard to reproduce the error...
>
> In any case DELETING THE FIRST LINES IN LJOFF ISN'T THE SOLUTION!!!
> I'm witting in capitals because it seems to work (no error message is
reported by leap), but no LJ corrections are applied!!!
>
> So the generated prmtop is wrong and useless because SIRAH depends on
those interactions to properly work. That explains the instabilities you
see during the simulations.
>
> There are some technical things about leap and the frcmod format, which I
won't go into detail now, but the number of blank lines is important to
properly read each section... I have a workaround but it needs more
testing...
>
> Hence, my STRONG advice at the moment is creating the topology files in
Linux or in a MacOS that doesn't give any error.
>
> 4) Coarse-grainig ligands is difficult due to the many relevant
interactions for binding. So far we don't have a systematic recipe to do it
and it really depends on the case... So I have no idea how you managed to
develop a compatible CG ligand for SIRAH...
>
> All the best,
>
> Matias
>
> ------------------------------------
> PhD.
> Researcher at Biomolecular Simulations Lab.
> Institut Pasteur de Montevideo | Uruguay
> [http://pasteur.uy/en/labs/biomolecular-simulations-laboratory]
> [http://www.sirahff.com]
>
> ----- Mensaje original -----
> De: "David A Case" <david.case.rutgers.edu>
> Para: "AMBER Mailing List" <amber.ambermd.org>
> CC: mmachado.pasteur.edu.uy
> Enviados: Viernes, 21 de Agosto 2020 23:06:09
> Asunto: Re: [AMBER] Questions (and a small bug report) about SIRAH in
Amber
>
> On Fri, Aug 21, 2020, Charo del Genio wrote:
>
> >
> >Anyway, the first question is: the Amber20 manual talks about
> >SIRAH, and even specifies where it should be found, i.e., in
> >${AMBERHOME}/dat/SIRAH. However... it isn't there, and in fact it isn't
> >even in the source tarball. Is it intended to be so?
>
> That is a bug in preparing the distribution files.  We'll get an update to
> fix this.  Thanks for the report.
>
> >
> >Second, upon downloading SIRAH and installing it in the above-mentioned
> >location, leap dies with a buffer overflow when sourcing
> >leaprc.sirah. The reason seems to be the presence of comment lines
> >within LJoff.frcmod, in the SIRAH directory.
>
> I can confirm the problem, but actual SIRAH users will have to chime in on
> how they get around this.
>
> [Matias: should we also have test cases in the Amber distribution?  Scott:
> would it be easy to skip comment lines (matching /^#/) in frcmod and
> parameter files?]
>
> This is Amber's first real foray into coarse-grained simulations, so there
> may well be some rough edges.  Thanks for helping out.
>
> ...thx...dac
>
> _______________________________________________
> 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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Aug 24 2020 - 09:30:03 PDT
Custom Search