Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Netiquette · Download · News · Gallery · Homepage · DSSR Manual · G-quadruplexes · DSSR-Jmol · DSSR-PyMOL · DSSR Licensing · Video Overview· RNA Covers

Messages - spreus

Pages: [1]
1
General discussions (Q&As) / Re: More than one helical region
« on: July 26, 2011, 08:59:43 am »
Hi Xiang-Jun

Thank you for clarifying. The ref_frames.dat file is exactly what I was looking for.

Best,
Søren

2
General discussions (Q&As) / More than one helical region
« on: July 25, 2011, 09:00:33 am »
Hi

I'm trying to rebuild pdb structures using the information optained from the 3DNA analysis output file. For structures containing more than one helical region, e.g. two different helices separated in space, I can't find a way to extract all the information regarding the relative position and orientation of the two helices relative to one another. I've attached the output file of pdb 1NNE as an example of two different helices separated by a kink between bp 11 and 12 - of which I would like to reconstruct the nucleic acid part.

The problem is that, as far as I can tell, the 3DNA output file only contains information about the origin (Ox, Oy, Oz) and the normal vector (Nx, Ny, Nz) of all base-pairs in the coordinate system of the given structure. However, the normal vector alone is insufficient to describe the exact orientation of the base pair which requires an additional axis in the bp ref frame to be specified, e.g. the local y-axis or the x-axis in addition to the bp z-axis.

My question then is: Is it possible using 3DNA to extract all the info required to rebuild a structure consisting of more than one helical region?

Best,
Søren

3
Hi Xiang-Jun

Ah okay, I think I get it now. For anyone else looking for the same information:
 B-DNA (generic) is the fiber model from Calf Thymus by Arnott & coworkers.
 B-DNA (generic alternate) is a fiber model of Premilat and Albiser.

References for the above models are provided in table 4 of the 2003 NAR paper on 3DNA.

Thanks,
Søren

4
The server seems to be overloaded quite often - I have experienced the same problems over the past several weeks (it is running extremely slow).
The server is highly useful so I hope it's possible to fix it running stable again.

Best,
Søren

5
I should probably specify my question better:
The "default" bp step parameter values I'm referring to are those used by the "B-DNA (generic)" rebuilding option in w3DNA.

All the best,
Søren

6
General discussions (Q&As) / Sequence dependent bp step parameters
« on: May 17, 2011, 01:18:07 pm »
Hi

I notice that when rebuilding B-DNA structures using w3DNA the default sequence dependent bp step parameters differ from what I can find in the DiProDB (http://diprodb.fli-leibniz.de/ShowTable.php).
So I was wondering if there is a reference for the default values used by 3DNA for the different dinucleotide steps?

Best,
Søren

7
General discussions (Q&As) / Re: Phi-angle problem?
« on: May 02, 2011, 06:00:08 am »
Hi again,

I found the problem. It was not in my calculation of Phi but in the calculation of T_g_ii. I see now it is supposed to be T_g_ii = T_g_i*T_i_ii (and not T_i_ii*T_g_i). Sorry if you have spent time looking into it.

I guess there is a typo in the SCHNArP paper equation 21 that confused me.

Best,
Søren

8
General discussions (Q&As) / Phi-angle problem?
« on: April 30, 2011, 01:36:10 pm »
Dear Xiang-Jun,

First of all, your work with 3DNA has been very inspiring and useful for me.

The problem:
I am currently trying to implement the CEHS scheme in a (Matlab) program of my own which I would like to be compatible with the base-pair step parameters output of a 3DNA-analysis. Following almost the exact matrix equations provided in the SCHNArP paper (J Mol Biol, 1997) and the original CEHS paper by El Hassan & Calladine (J Mol Biol, 1995) I've made a script that rebuilds a structure based on bp step parameters as the input. The program produces a .mol file so I can check the output structure with the output structure from the 3DNA-rebuilding option using the same input parameters.
However, it now appears that the produced structure is identical to the corresponding 3DNA output only for small roll/tilt values, whereas for large roll/tilt it is not. The deviation seems to be depending on the position in the sequence: E.g. using the input parameters provided in the attached file the 3DNA/my_script structures virtually deviates only at the last three base-pairs. Then, making only the last three base-pairs (bottom in attached file) the relative orientation and position of these three base pairs is not the same as in the 10 bp structure, despite their local bp step parameters are set to be the same.

The question:
Since the problem seems to be related to roll/tilt I'm suspecting it lies in my calculation of the angle 'Phi'. The value I'm refering to is the 'Phi' in equation (18-20) in the SCHNArP paper and also in equation (1) in the 3DNA paper (NAR, 2003). Therefore, I hope the answer to the following question might solve my problem:
- How exactly does SCHNArP and 3DNA calculate the 'Phi'-angle of each bp-step when rebuilding a structure based on shift/slide/rise/tilt/roll/twist inputs? Does it use an identity matrix to represent the MST at every base pair step or just the first one?

I hope you have time to help with this despite it's not directly related to the use of 3DNA.

All the best,
Søren

PS. If it is of any help, this is the relevant piece of my Matlab code:

Code: [Select]
...

 % Make first bp:
    r_i = [0 0 0];
    T_i = eye(3,3);
    MST = eye(3,3);
    XYZ(:,1-3) = xyz(:,1-3); % Here XYZ is coordinates in global ref. frame, xyz is in the local bp ref. frame.

 % Make consecutive bp's:
for j = 2:NoOfBasePairs
  ***(it first calls input parameters and coordinates in the bp ref. frames). Then:***

    RollTilt = sqrt(roll^2+tilt^2);
    rolltiltaxis = normr((tilt/RollTilt)*MST(:,1)'+(roll/RollTilt)*MST(:,2)');

    Phi = acos(dot(rolltiltaxis,MST(:,2)));
     if dot(cross(rolltiltaxis,MST(:,2)),MST(:,3)) > 0
         Phi = sqrt(Phi^2);
     elseif dot(cross(rolltiltaxis,MST(:,2)),MST(:,3)) < 0
         Phi = -sqrt(Phi^2);
     end

     % Make T matrix and translation vector:
     T_i_mst = Rz(twist/2-Phi)*Ry(RollTilt/2)*Rz(Phi);
     T_i_ii = Rz(twist/2-Phi)*Ry(RollTilt)*Rz(twist/2+Phi);
     r_i_ii = [shift slide rise]*T_i_mst';
     
     T_g_ii = T_i_ii*T_g_i;
     r_g_ii = r_g_i+r_i_ii*T_g_ii';
     
    % Rotation and translation of base pair ii:
     run = xyz(:,(j*2+j-2):(j*2+j))*T_g_ii';
     XYZ(:,(j*2+j-2):(j*2+j)) = [run(:,1)+r_g_ii(1) run(:,2)+r_g_ii(2) run(:,3)+r_g_ii(3)];
   
     T_g_i = T_g_ii;
     r_g_i = r_g_ii;
   
end

% Then, visualization of coordinates in XYZ
...

Pages: [1]

Created and maintained by Dr. Xiang-Jun Lu [律祥俊] (xiangjun@x3dna.org)
The Bussemaker Laboratory at the Department of Biological Sciences, Columbia University.