RE: AMBER: Is the MD simulation normal

From: Yong Duan <>
Date: Thu, 26 May 2005 10:05:14 -0700

Dear Shulin,

You perhaps can learn more by looking at,

1) How soon did you loss that hydrogen bond and did it ever go back.
2) Is the hydrogen bond exposed to solvent? This can give you clue as to
the stability.
3) It sounds like that your complex is a bit unstable. Any experimental
data on binding affinity (IC50 or Kd)?

Keep in mind, hydrogen bonds are weak interactions. Lossing one or two
is quite common even for stable complex. Hydrogen bonds of 3A or so are
not considered stable, if that is the distance your system started from.

Last, the notion of importance in binding can be viewed in many ways.
One way is that it helps stabilize the complex and the other way is that
it enhances specificity (without contributing much to the affinity). For
hydrogen bonds, the roles are mainly in the second aspect. Afterall,
those atoms that form hydrogen bonds in complex can be just as happy (or
even happier) to form hydrogen bonds with water. There are, mind you, a
lot of water molecules around in water and each has the ability to form
four hydrogen bonds.

Hope this helps ...


-----Original Message-----
From: [] On Behalf
Of Shulin Zhuang
Sent: Thursday, May 26, 2005 8:31 AM
Subject: AMBER: Is the MD simulation normal

Dear everybody,

Now, I have a very troublesome problem. I have performed one MD
simulation on a protein-inhibitor complex with its X-crystal structure
as the initial structure. In the X-crystal structure, the inhibitor has
two important hydrogen bonds with the protein. For each of two hydrogen
bonds, the distance between donor and receptor is within 3 angstorm.

I use ff03 force field to carry out the MD simulation with Amber8.
The partial atomic charge for the ligand was generated using RESP
methods. I use constraint minimization and constraint NVT MD of 50 ps
to heat up the solvated system.

After the minimization and NVT MD, I get the complex structures using
command ambpdb with the generated restrt file and find that after the
minimization stage or after the NVT MD stage, the two important hydrogen
bonds still exsit.

However, After 1 ns NPT MD, the two important hydrogen bonds were lost.
I roughly measure the -- the distance between donor and receptor , the
distance become 2 or 3 angstrom larger than the former distance.

Is my NPT MD normal? If it is wrong, where is the problem.

Looking forward to the answer!

Best regards!

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Thu May 26 2005 - 18:53:00 PDT
Custom Search