The program quadric_diffusion uses the quadratic representation approach
of Brüschweiler, et al. [Science 268:886-889, (1995)] and
Lee, et al. [J. Biomol. NMR 9:287-298 (1997)] to determine
the diffusion tensors for spherical, axially-symmetric and fully anisotropic
motional models from experimental nitrogen-15 or carbon-13 spin relaxation data.
The citation for quadric_diffusion is:
L. K. Lee, M. Rance, W. J. Chazin and A. G. Palmer, Rotational diffusion anisotropy of proteins from simultaneous analysis of 15N and 13Cα nuclear spin relaxation, J. Biomol. NMR 9, 287-298 (1997).
Usage: quadric_diffusion control_file
axial_ratio_low axial_ratio_high #steps
number_of_atom_types 'atom_1' ... 'atom_n'
atom_1_tm_file_name
...
atom_n_tm_file_name
input_pdb_file_name
output_pdb_file_name (for axially symmetric model)
output_pdb_file_name (for anisotropic model)
The first line constains estimates of the minimum and maximum axial
ratio (Dpar/Dper) for grid searching, and the number of steps to use for
the grid search.
Each atom name must be enclosed in single quotations, e.g. 'N' or 'CA'
for amide nitrogen or alpha carbon spins. For example, to analyze nitrogen-15
data, the second entry would be:
1 'N'
To analyze nitrogen-15 and carbon-13 data simultaneously,
the second entry would be:
2 'N' 'CA'
The atom_n_tm_file contains the local correlation time of atom_n for each residue determined either from a model-free analysis of nitrogen-15 or carbon-13 spin relaxation data using the modelfree program or determined from the nitrogen-15 or carbon-13 R2/R1 ratio using the r2r1_tm program described below. One file must exist for each atom type. The atom_n_tm_file contains one line for each residue to be analyzed. Each line contains the following fields:
residue_number tm tm_uncertainty
For both the control_file and tm_file, lines starting with "#"
are ignored. The residue number in the tm_file must match the residue number
in the input_pdb_file.
For each model (spherical, axially-symmetric, and anisotropic diffusion),
the diffusion tensor and Euler angles are output. The chi-square and F statistics
are calculated for comparing models. The program also returns the predicted
effective diffusion constant for each residue and the squared residual from
the fitting process. For diagnostic purposes, spherical harmonic functions
of the NH or CH bond vectors (in the diffusion reference frame) are also output.
For the anisotropic model, uncertainties are obtained by Monte Carlo simulationsand the results for uncertainties for the phi and psi angles can be inaccurate
if the values of the angles are near 0, pi or -pi.
The program contains a crude fix,
but if your values of phi or psi are near these values you might want to print
the actual Monte Carlo results and inspect them. See the program listing for more information.
For axially symmetric and anisotropic diffusion models, the structure in
the input_pdb_file is rotated to the diffusion frame. To be meaningful,
prior to analysis, the center of mass of the molecular structure should
be translated to coordinate origin by using the program pdbinertia.
The angles phi, theta, and psi represent the Euler angles in the z-y-z convention with the
rotations being applied to the original molecule to generate the rotated molecule. In the z-y-z
convention, the first rotation is about z, the second rotation is about the intermediate y axis and
the third rotation is about the final z-axis.
Output structures from the two analyses may differ by 90° rotations.
The program r2r1_tm calculates a local effective correlation time
from the R2/R1 ratio, in which R2 and R1 are the spin-spin and spin-lattice
relaxation rate constants for nitrogen-15 or carbon-13 spins. The local
correlation times are used as input for the quadric_diffusion
program.
Usage: r2r1_tm [-1] > tm_file
The program prompts for the following information:
nucleus_type (15N or 13C)
XH_bond_length (Ang.) X_CSA (ppm) [enter a blank line to accept defaults]
proton_spectrometer_field for R1 (MHz)
proton_spectrometer_field for R2 (MHz)
initial_tm_estimate (ns)
number_of_entries (in the r2r1_file)
r2r1_file_name
The input r2r1_file contains one line for each residue to be analyzed. Each line contains the following fields:
residue_number r2/r1_ratio r2/r1_uncertainty
The input residue numbers should correspond to residue numbers in a pdb
file for the molecule. Lines beginning with "#" are ignored in
the r2r1_file. If the -1 flag is included on the command line, then
the input r2r1_file is not read and the user is prompted to enter a single
r2/r1_ratio and uncertainty. The functionality with the -1 flag is essentially
the same as using the 'tmest' program of earlier releases.
The output tm_file contains one line for each residue analyzed. Each line
contains the following fields:
residue_number tm tm_uncertainty