AW: problem with water molecules angle in MD simulation

From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Wed Jan 28 2015 - 04:08:53 CST

So far the file looks ok compared to mine, only the order of the force constants and equil values are different, but correctly inserted into the bonds_inc_hydrogen table. Can I see your namd input file?

 

Norman Geist.

 

Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im Auftrag von Shoresh
Gesendet: Dienstag, 27. Januar 2015 17:31
An: namd-l_at_ks.uiuc.edu; Norman Geist
Betreff: Re: namd-l: problem with water molecules angle in MD simulation

 

Hi Norman,
Thanks for all the useful hints you have provided.

The parameter file for one water molecule is attached.

Best,

Shoresh

 

 

On Tue, Jan 27, 2015 at 3:28 AM, Norman Geist <norman.geist_at_uni-greifswald.de> wrote:

Hi,

 

actually I meant that you create a parm file with only one water molecule, so I can compare it to my files.

 

You are right, the three bonds should fix the angle. You must also know that the water models and so SHAKE (rigidbonds) are hard coded in NAMD. This means if NAMD is somehow unable to detect your water as water, it won’t work. NAMD does this detection based on the atoms masses. This means you may not change them.

 

The point about minimization might be correct, but I’m sure NAMD doesn’t turn of SHAKE during minimization.

 

Send me a parmfile with only one water. I’ve observed weird behavior earlier when the leap.rc isn’t properly set up.

 

Norman Geist.

 

Von: Shoresh Shafei [mailto:s.shafei_at_gmail.com]
Gesendet: Montag, 26. Januar 2015 17:54
An: Norman Geist
Betreff: Re: namd-l: problem with water molecules angle in MD simulation

 

Dear Norman,

Thank you for your prompt response.

 

As you mentioned, TIP3P does not have angle property, however, as far as I remember two OH bonds are determined. Doesn’t fixing "three bonds" automatically fix the angle?

 

I was not sure what you meant by onewater.parm7 file so I directly replied to your own email address.

All I could think of was parm10.dat that I use in Amber which is attached to this email.

 

I really appreciate your help.

Best

Shoresh

 

 

 

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:20:51 CST