Re: [AMBER] Problems on IRED analyses for order parameter

From: David A Case <>
Date: Sun, 12 Jan 2014 13:52:24 -0500

On Sun, Jan 12, 2014, Ye Wei wrote:
> I am currently doing IRED analyses to calculate S2 (order parameters).
> I am using Lysozyme as a test example.
> Following is my Cpptraj Script:
> "
> trajin
> vector v2 :2.N ired :2.H
> vector v3 :3.N ired :3.H
> ...
> vector v126 :129.N ired :129.H

Make sure you have exactly 126 vectors above; I don't know what is in the
"...", but it looks like you have vectors v2 to v126, which would be 125.

> matrix ired name matired order 2
> diagmatrix matired vecs 126 out ired.vec name ired_vec
> ired relax freq 500 order 2 tstep 200.0 tcorr 100000.0 out output.out noefile noe modes ired_vec
> "
> Some reference papers suggest the sampling time window should be 200 ps,
> so I set tstep=200

No: tstep should be the actual time step between your trajectory frames. It
has nothing to do with what anything (including the "sampling time window")
*should* be.

> And the 100000.0 shows my MD traj is 100 ns' long.

No: the tcorr time is the time over which to caclulated correlation functions,
which should generally be much shorter than the length of the simulation.

> I am so confused about the description about tstep and tcorr in
> AmberTools Manual, what is their unit? [ps/ns]?

You can choose what units you like, as long as tstep and tcorr are the same.
(I'd use ps to be a little safer, however. Order parameters don't depend on
the time unit, but things like relaxation rates do.)

See the example/test calculations in

.../hope this helps...dac

AMBER mailing list
Received on Sun Jan 12 2014 - 11:00:02 PST
Custom Search