# Re: [AMBER] [Amber] nmode() results with imaginary frequencies

From: Jason Swails <jason.swails.gmail.com>
Date: Tue, 1 Dec 2015 22:05:55 -0500

On Tue, Dec 1, 2015 at 6:18 PM, Guqin Shi <shi.293.osu.edu> wrote:

> Hi all,
>
> I'm using nmode() to calculate entropy for protein-ligand complex. I got
> results successfully but some of them has additional line seen in below:
> ==========================================================
> - Thermochemistry -
>
> Temperature: 298.150
> Pressure: 1.000
> Mass: 11668.480
> Principal moments of inertia in amu-A**2:
> 809809.49 1666184.08 1773921.38
> Rotational symmetry number is 1
> Assuming classical behavior for rotation
> Rotational temperatures: 0.000 0.000 0.000
> *found 1 imaginary frequencies*
> Zero-point vibrational energy: 8667.328
>
> freq. E Cv S
> cm**-1 kcal/mol cal/mol-K cal/mol-K
> Total: 6749.690 3252.715 3898.857
> translational: 0.888 2.979 53.869
> rotational: 0.888 2.979 53.610
> vibrational: 9204.459 3246.757 3791.379
>
> ==========================================================
>
> Is this a big problem I need to solve before I use the data? I read a
> little bit it seems that it indicates the protein is in some sort of
> transition state...? I used xmin() to minimize the protein to grms of
> 0.001. I also browsed amber mailing list and in a post from 2009, somebody
> suggested that ntrun() set to 0 could solve problem. My nmode() function is
> set as:
> nmode(xyz,3*natm,mme2,0,0,0.0,0.0,0); where ntrun() is already set to 0.0.
>
> Any idea or something important I missed?
> Thanks a lot for your time!
>

​Check the actual printed frequencies. Remember that 6 frequencies
*should* be zero -- corresponding to the 3 translational and 3 rotational
degrees of freedom. If these eigenvalues are negative (i.e., an imaginary
frequency), rather than 0, this doesn't really matter too much since those
modes are omitted anyway. However, if the 7th mode is imaginary, you
probably need a stricter minimization tolerance.​

HTH,
Jason

```--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
```
Received on Tue Dec 01 2015 - 19:30:04 PST
Custom Search