Re: Protein collpasing problem!

From: Axel Kohlmeyer (
Date: Mon Jun 29 2009 - 08:25:06 CDT

On Mon, 2009-06-29 at 14:12 +0800, ydhuang2727 wrote:

dear yangdong,

> Hi,
> Yes, you are right, i should not have directly heated the protein up
> to 310K, which may cause problems. And i should heat gradually, to
> make the whole system approximately equilibrium. But from the namd
> tutorial, i did not find the more standard procedure, as you

please take under consideration, that the namd tutorial is meant
to show how to use the namd program, but not necessarily the best
practices for MD simulations, which in turn depend _very_ much
on the system at hand. the ubiquitin example used in the tutorial
is very uncritical and the tutorial itself is optimized to get
people _quickly_ through the whole process of setting up, running
and analyzing a trajectory without cutting too many corners.

setting up a larger and more delicate system, may require some
or sometimes even a lot more care and a lot of common sense.
you have shown common sense in wondering that your structure
is not preserved, now you should think about what could have
all gone wrong and try to make up for it.

> suggested, involving slowly heating system to the desired
> temperature. And i also feel surprise that you are so surprise,
> because i did not found anything wrong using "minimizing and then
> jumping to 310K", which is also definitely referred to in the namd
> tutorial.
> I myself previously used a loop in the .conf file to heat the protein,
> which refers to increase the heat source temperature up to desired
> temperature 20K by 20K that does look proper way but takes time, you
> know. Since heating directly up to 310K got the same equilibrium state
> as step by step, i chose the former to save cost.

people generally underestimate the time it needs to equilibrate
a system. furthermore, just gently heating up is not the only
protocol to bring a system to the desired state, and if you
heat up the system by rescaling, you should have a careful look
at the protein itself. quite often the protein is at a much
higher temperature than the surrounding water.

> As your suggested standard procedure, i do have never seen anything
> about it.
> Will you please offer us some information available.

many groups or people have their own protocol to bring a system
close to equilibrium. there is not one standard procedure. you
always have to keep the specific properties of your system in
mind and potential "hot spots". if you are routinely cutting
corners to save time, you better double check, if this is really
working. it might actually cost you much more time than being
more careful.
> Have you have any good advice on this problem.

use common sense. don't follow _any_ advice blindly. always
prove to yourself that you are making the right choices.
experiment with simple systems and be aware that a smaller/simpler
system is always more forgiving than a large system.

> Good day!
> Yandong,
> ______________________________________________________________________
> 200万种商品,最低价格,疯狂诱惑你

Axel Kohlmeyer
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
If you make something idiot-proof, the universe creates a better idiot.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:58 CST