From owner-chemistry -A_T- ccl.net Thu May 18 22:09:00 2006 From: "Nathan Scott scottjn-,-gmail.com" To: CCL Subject: CCL: QM/MM computational cost? Message-Id: <-31786-060518215707-30329-R8w/DBiVEzp/ly4wIDIJTw%a%server.ccl.net> X-Original-From: "Nathan Scott" Content-Type: multipart/alternative; boundary="----=_Part_103073_3330678.1148003815778" Date: Thu, 18 May 2006 21:56:55 -0400 MIME-Version: 1.0 Sent to CCL by: "Nathan Scott" [scottjn(_)gmail.com] ------=_Part_103073_3330678.1148003815778 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Thanks to everyone for their help! I appreciate it very much. Ross, thank= s for the insight, as well as the time cost data. You mention that a full ab initio treatment "could get very expensive." I was actually planning to us= e B3LYP/6-31G(d,p), and to do normal mode analysis of the QM system at a few different points in the trajectory. Could you (or anyone else) take a gues= s at how much slower your calculations would have been had you used B3LYP/6-31G(d,p) rather than PM3? Am I right in thinking that the normal mode analysis of the QM system is going to be extremely expensive? J. Nathan Scott Graduate Student University of Pennsylvania Biochemistry and Molecular Biophysics Graduate Group Work Phone: +1.215.898.8783 On 5/18/06, Ross Walker ross]|[rosswalker.co.uk wrote: > > Sent to CCL by: "Ross Walker" [ross|*|rosswalker.co.uk] > Hi Nathan, > > > Greetings CCLers. Can anyone point me to a reference discussing the > > computational cost of QM/MM simulations and how it scales > > with size of the > > components? I have some experience with quantum chemistry > > simulations, as > > well as molecular dynamics simulations, but I'm new to QM/MM > > and am being > > asked as part of a proposal I'm writing to discuss the > > computational cost o=3D > > f > > the calculations I want to do. Right now the plan is to > > solvate a simple 6 > > atom quantum mechanical system with approximately 1000 TIP4P water > > molecules. > > This really depends on the type of QM calculation you plan on doing which > you don't state. Semi-empirical is significantly cheaper than full ab > initio. For what you suggest above if you wanted to do semi-empirical you > can use Amber 9 and the computational cost will be about 6% more than > doing > the calculation classically. You can probably go to about 90 QM atoms or > so > before it even starts to become expensive. Plus you can do a simulation > with > a full treatment of electrostatics and periodic boundaries for both the Q= M > and MM regions using PME. Without a PME treatment you will need to use an > infinite (i.e. no) cutoff and probably add an implicit solvent treatment > for > the gas phase region of your system to get meaningful results. Using a cu= t > off with gas phase / water cap simulations introduces serious artefacts > into > your simulation and can completely invalidate your results. Hence often i= t > can be quicker to do a full PME treatment than to run the gas phase > non-periodic simulation. > > A lot of work has been put into making Amber 9's QM support as fast as > possible while preserving accuracy in the gradients. It is around 10 time= s > quicker than Amber 8 and Charmm c32. The Amber 9 QMMM implementation is > currently being added to Charmm and will I believe be released shortly. > > As a quick guide here are the timings I got for two water molecules > treated > with PM3 and then solvated with a rectangular box consisting of 332 TIP4P > water molecules. PME was run with an 8 angstrom MM and QM direct space cu= t > off, shake was used for both the MM and QM region, langevin thermostating > was used at 300K and a timestep of 2fs was used. For the gas phase > simulation I used an infinite MM and QM cut off. The default SCF > convergence > of 1.0x10-8 KCal/mol was used. > > This is for a pentium-D 3.4GHz 2MB L2 cache machine. Time is for 2000 > steps > =3D 4ps. > > 1 cpu 2 cpu's > time(s) ns/day time(s) ns/day > PME classical =3D 39.99 8.64 23.35 14.80 > PME QMMM =3D 62.15 5.56 40.15 8.83 > Gas phase clas=3D 123.67 2.79 64.48 5.36 > Gas phase QMMM=3D 130.99 2.64 70.12 4.93 > > So for the system you want to study it really won't cost you much at all. > However, if you want to do a full abinitio treatment then it could get > very > expensive. > ------=_Part_103073_3330678.1148003815778 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Thanks to everyone for their help!  I appreciate it very much.  R= oss, thanks for the insight, as well as the time cost data.  You menti= on that a full ab initio treatment "could get very expensive."&nb= sp; I was actually planning to use B3LYP/6-31G(d,p), and to do normal mode = analysis of the QM system at a few different points in the trajectory. = ; Could you (or anyone else) take a guess at how much slower your calculati= ons would have been had you used B3LYP/6-31G(d,p) rather than PM3? &nb= sp; Am I right in thinking that the normal mode analysis of the QM system i= s going to be extremely expensive?

J. Nathan Scott
Graduate Student
University of PennsylvaniaBiochemistry and Molecular Biophysics Graduate Group
Work Phone: +1.21= 5.898.8783


On 5/18/06, Ross Walker ross]|[rosswalker.co.uk= <owner-chemistry]=[ccl.net= > wrote:
Sent to CCL by: "Ross Walker" [ross|*|rosswalker.co.uk]
Hi Nat= han,

> Greetings CCLers.  Can anyone point me to a refe= rence discussing the
> computational cost of QM/MM simulations and ho= w it scales
> with size of the
> components?  I have some experi= ence with quantum chemistry
> simulations, as
> well as molecul= ar dynamics simulations, but I'm new to QM/MM
> and am being
> = asked as part of a proposal I'm writing to discuss the
> computational cost o=3D
> f
> the calculations I want = to do.  Right now the plan is to
> solvate a simple 6
&g= t; atom quantum mechanical system with approximately 1000 TIP4P water
&g= t; molecules.

This really depends on the type of QM calculation you plan on doing= which
you don't state. Semi-empirical is significantly cheaper than ful= l ab
initio. For what you suggest above if you wanted to do semi-empiric= al you
can use Amber 9 and the computational cost will be about 6% more than d= oing
the calculation classically. You can probably go to about 90 QM ato= ms or so
before it even starts to become expensive. Plus you can do a si= mulation with
a full treatment of electrostatics and periodic boundaries for both the= QM
and MM regions using PME. Without a PME treatment you will need to u= se an
infinite (i.e. no) cutoff and probably add an implicit solvent tre= atment for
the gas phase region of your system to get meaningful results. Using a = cut
off with gas phase / water cap simulations introduces serious artefa= cts into
your simulation and can completely invalidate your results. Hen= ce often it
can be quicker to do a full PME treatment than to run the gas phase
= non-periodic simulation.

A lot of work has been put into making Ambe= r 9's QM support as fast as
possible while preserving accuracy in the gr= adients. It is around 10 times
quicker than Amber 8 and Charmm c32. The Amber 9 QMMM implementation is=
currently being added to Charmm and will I believe be released shortly.=

As a quick guide here are the timings I got for two water molecules= treated
with PM3 and then solvated with a rectangular box consisting of 332 TIP= 4P
water molecules. PME was run with an 8 angstrom MM and QM direct spac= e cut
off, shake was used for both the MM and QM region, langevin thermo= stating
was used at 300K and a timestep of 2fs was used. For the gas phase
s= imulation I used an infinite MM and QM cut off. The default SCF convergence=
of 1.0x10-8 KCal/mol was used.

This is for a pentium-D 3.4GHz 2MB L2 cache machine. Time is for 2000 steps
=3D 4ps.

 &nbs= p;            &= nbsp;      1 cpu     &nbs= p;         2 cpu's
  &= nbsp;           &nbs= p;  time(s) ns/day      time(s) ns/dayPME classical =3D   39.99  8.64   &nbs= p;    23.35   14.80
PME QMMM      =3D   62.15 &nbs= p;5.56        40.15  &nbs= p; 8.83
Gas phase clas=3D  123.67  2.79 &n= bsp;      64.48    5.36Gas phase QMMM=3D  130.99  2.64   &nbs= p;    70.12    4.93

So for t= he system you want to study it really won't cost you much at all.
However, if you want to do a full abinitio treatment then it could get = very
expensive.

------=_Part_103073_3330678.1148003815778--