IUBio GIL .. BIOSCI/Bionet News .. Biosequences .. Software .. FTP

Questions regarding relaxation matrix refinement

Rasmus Storjohann rstorjoh at sfu.ca
Thu May 28 22:01:07 EST 1998

I have a couple of questions:

1) I have a restraint file containing build-up curves for various NOE 
interactions, and with varying number of data points on each buildup 
curve, e.g.

Class 300
Class 250
assign () () 23000 20000
Class 300
Class 250
Class 200
assign () () 54000 40000 35000

At the second Class 300 and Class 250 statements I get this warning:

%RELAX-WRN: Class already defined

Is it safe to ignore this warning, or is it something I should worry about?

2) I want to use NOE data acquired between 75ms and 300ms mixing times. 
According to the taugrid.inp routine, my correlation time (isotropic 
motion, no internal motion) is 10 ns. What value should I use for the 
relaxation cutoff? The X-plor manual suggests that 4.5A is an appropriate 
value for tau = 2.3ns and mixing time = 200ms, so I know I will have to 
use a higher value than that, but I don't know how much higher. What is 
the justification of using a relative vs. an absolute threshold value.

3) When I run dgsa.inp and refine.inp the molecular dynamics runs fine 
with a time step of 0.005 ps. The relaxation matrix refinement crashes 
unless I use 0.001 ps. Why is that, and is there any way around this 

4) The relaxation matrix refinement script at the X-plor web site uses a
"parabola" potential energy function for the relaxation matrix
pseudo-energy term rather than the "well" potential energy function. What
is the reason for this choice? I have some spectral regions whihc are
quite crowded, and I want to use a higher error estimate of NOE
measurements from these regions than of NOEs taken from well resolved
regions of the spectrum. To do this, I need to use the "well" potential
energy function, and I want to know if there is any reason I ought to 
stick to the "parabola" option.

5) Is there a way I can get X-plor to ignore NOE intensities at a 
particular mixing time, i.e. to exclude one class of NOEs from the 
calculations. Setting the weight to zero did not seem to help (I did that 
for all the classes of NOEs, but the calculation didn't get any faster).

6) How much does the CPU time requirement depend on how many different 
mixing times are included? 

Thank you for your patience,

        __		Rasmus Storjohann
       / /\		Institute of Molecular Biology
      / /  \		and Biochemistry
     / / /\ \		Simon Fraser University
    / / /\ \ \		Burnaby, British Columbia V5A 1S6
   / /_/__\ \ \		e-mail: rstorjoh at sfu.ca
  /________\ \ \	Phone: (604) 291-5657 / 415-0575
  \___________\/	FAX:   (604) 291-3765

More information about the X-plor mailing list

Send comments to us at archive@iubioarchive.bio.net