Re: [AMBER] Matrix correlation

From: Sylvester Tumusiime <>
Date: Sun, 4 Jan 2015 23:58:12 +0000

Hello Dan,

Thank you very much for you advice. I found it to be very helpful.

From: Daniel Roe []
Sent: Friday, January 02, 2015 12:46 PM
To: AMBER Mailing List
Subject: Re: [AMBER] Matrix correlation


On Thu, Jan 1, 2015 at 5:19 PM, Sylvester Tumusiime <> wrote:
> 1. Does it matter whether I place the reference file before I trajin my netcdf file or do I need to place it after I trajin my netcdf file? ( I plan to use the data in the netcdf file for the correlation analysis)

No, the order of 'trajin'/'reference' doesn't matter since they are
separate types of input data. What will matter is the order of a
'reference' command and any action that will use that reference
structure, so e.g. your 'reference' command needs to come before the
'rms' command.

> For example below is the input i plan to use for my analysis:

I will provide some notes on your input below:

> reference flp.inpcrd
> trajin flp.netcdf 1 1000000000 1
> rms TOFIRST :1&!.H= first out flp.agr mass
> autoimage
> center :1 origin mass

Never *never* image after rms fitting, since the rms fit will almost
certainly rotate your coordinates out of alignment with your original
unit cell vectors. In fact, recent versions of cpptraj print a warning
to this effect:

Warning: Coordinates are being rotated and box coordinates are present.
Warning: Unit cell vectors are NOT rotated; imaging will not be possible
Warning: after the RMS-fit is performed.

If you really need to image it needs to come *before* rms fitting.
However since in this case you're just calculating a correlation
matrix for a single molecule I don't think you even need to autoimage.
In fact it's probably better to just strip out all of your solvent
with a 'strip' command prior to the rms fit. Also, since you're just
rms fitting to the first frame you don't actually need your reference
structure here.

> matrix correl name flp out Flp.dat .1 by atom

First, is this the right mask? Are you sure you wanted to calculate
the correlation of a single atom? Second, the keyword is 'byatom' (all
one word). This input shouldn't even work with recent versions of
cpptraj (it will exit because of the malformed keyword), so make sure
you're using the latest version (14.22).

> 2. Can i follow this with:
> - analyze matrix
> and
> - analyze modes
> to get the necessary data?

Depends on what 'necessary data' you're looking for. It may be enough
for you to change 'Flp.dat' to 'Flp.gnu' and visualize the correlation
matrix in gnuplot.

Hope this helps,


> Thank you,
> Silver
> _______________________________________________
> AMBER mailing list

Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 307
Salt Lake City, UT 84112-5820
(801) 587-9652
(801) 585-6208 (Fax)
AMBER mailing list
AMBER mailing list
Received on Sun Jan 04 2015 - 16:00:02 PST
Custom Search