Re: amber-developers: fixup to max14

From: David A. Case <>
Date: Wed, 12 Nov 2008 17:55:43 -0500

On Wed, Nov 12, 2008, Lachele Foley wrote:
> Yes, and no. Since using ntb=1 gave me a "vacuum rift" ("vacuum bubble"
> hardly seems a sufficient term), I tried the minimization with ntb=0.
> And... I can get it to run using the serial version or using the MPI
> version with only one processor. But, with 2 or 4 processors the run hangs.
> No complaint. Just hangs. The different variants are copied below.
> Here is a summary of run behavior. This uses the same top & crd files I've
> sent before.

I think the basic problem is that Mike tried to trick his periodic list
builder to work with non-periodic problems, and there are probably weird
things there (and maybe with periodic list builds as well?) that assume
certain things about the density of atoms. Systems with lots of extra points
may violate those assumptions, and not get properly trapped.

Some questions:

1. Why are you doing lots of minimization in the first place? The usual idea
(with explicit solvent) is to only do minimization if the starting structure
is so bad that MD will fail. 14,000 cycles seems way beyond what is necessary
or productive.

2. Related: why do ntb=0 if you have explicit solvent?

3. If I read correctly, you are starting with a very low density of water (0.6
g/cm3)? It is hardly surprising that fully minimizing this would lead to
water bubbles or rifts. What you want to do is to get going with MD, get the
system heated to some non-zero temperature, then set ntp=1 and get the system
moving to a reasonable density.

4. You seem not to be using pmemd. Of course, this is good in that it may
expose sander bugs, but Bob's code is a lot less likely to have weird corners
that mysteriously hang, and it will almost always be faster to boot. After we
twisted his arm to get extra points into pmemd, we should take advantage of

> Yeah... so I think the calculation is right, but I'm a little puzzled by
> the starting structure. I just had leap solvate a water molecule using
> "solvatebox m TP5 10". The resulting box has a density of 0.622 g/cc,
> according to leap.

(a) use the option closeness parameter (say set to 0.7) to get a little more
reasonable initial density;

(b) we eventually need a TIP5PBOX setup (rather than just solvating with TP5).

I hope this all works--I want to find out what sucrose does with TIP5P and the
corresponding glycam potentials!

Received on Fri Dec 05 2008 - 14:35:54 PST
Custom Search