Messages

1476

General discussions (Q&As) / Re: trajectory analysis: error with analyze

« on: August 05, 2009, 08:50:16 pm »

I figured out an error with my simulations - I hadn't taken care of PBC. So, after fixing it I no longer get the error I had mentioned.

However, I would still like to know a bit more about the error.

I am glad that you figured out where the problem was.

Regarding the meaning of the error message,

Code: [Select]

some residue has more than one pair
residue index 24 too big (> 22)
please check your input fileIt would help clarify the issue greatly if you include a minimal, producible example so others can understand exactly what you mean. Broadly speaking, it means the input file to analyze does not match the corresponding PDB file. More specifically, in your input file, a nucleotide is specified with sequential no. 24, while the PDB has only 22 residues.

HTH,

Xiang-Jun

1477

General discussions (Q&As) / Re: 3DNA version 2.0

« on: July 27, 2009, 07:46:17 pm »

Thanks for your interest in 3DNA and for sharing your difficult experience with the user community. Over the past few months, I have received (being CC-ed) dozens of email requests for 3DNA v2.0, and I am just wondering if anyone has a more positive story to cheer up the mode of this thread a bit -- any feedback?

Technically, I have already done everything as far as 3DNA v2.0 itself is concerned. What's leftover is distribution, the final stage of a software product, presumably an exciting experience both to the author and the user community. Rutgers has licensed 3DNA and required that it is distributed in a protected fashion, with user registrations etc. In principle, it is easy to release the tarballs of 3DNA v2.0, which I compiled early this year, automatically through a web-form. Even dealing with each individual user's request manually, on a case-by-case basis, should take no longer than a few minutes. Over the times, I have mentioned to Rutgers several times that distributing 3DNA as widely as possible (with proper measure of protection of intellectual property) is to the benefit of everyone who really cares about the software product, and it is also an obligation per the 2008 3DNA Nature Protocols publication.

Xiang-Jun

1478

General discussions (Q&As) / Re: 3DNA version 2.0

« on: July 20, 2009, 07:28:36 pm »

I have registered in 3DNA forum. Would you please give details about downloading 3DNA 2.0 (i read in 3DNA forum that it requires some license, password etc)

Please note that registration with the 3DNA forum is unconnected (at least for the time being) with downloading 3DNA v2.0. I have tried hard to make my point simple and unambiguous, so please follow instructions there.

Xiang-Jun

1479

General discussions (Q&As) / Re: DNA Rotation

« on: July 08, 2009, 11:50:47 pm »

Thanks for using 3DNA. I am glad that you are taking advantage the visualization part of 3DNA: the other day, I blogged on blocview wondering why it has not been used that much.

Could you be more specific? Which command and options did you use? Preferably with a reproducible example?

Xiang-Jun

1480

General discussions (Q&As) / Re: defining a local helix axis

« on: July 08, 2009, 11:15:55 pm »

Thanks for using 3DNA, and posting interesting questions on defining local helix axis and quantifying the kinks in a DNA structure.

3DNA outputs many structural parameters including "Origin (Ox, Oy, Oz) and mean normal vector (Nx, Ny, Nz) of each base-pair in the coordinate system of the given structure", and "Position (Px, Py, Pz) and local helical axis vector (Hx, Hy, Hz) for each dinucleotide step", as shown in file bdl084.out distributed with 3DNA. In idealized cases, e.g., where a perfectly regular B-DNA fragment connects to a perfectly regular A-DNA fragment, the local helical axis vector should do the trick. However, in normal cases, e.g., for your protein-bound DNA, local distortions make local helical axis associated with each dinucleotide quite irregular (see an worked example below), and as always, one needs to be careful in drawing conclusions from using it.

Local kinks in protein-bound DNAs are normally quantified using roll angle in the literature. See, for examples, the two Chen et al. papers: http://www.ncbi.nlm.nih.gov/pubmed/11724532 and http://www.ncbi.nlm.nih.gov/pubmed/11724533.

Also, you may want to check Curves, which has been used frequently in the literature for quantifying DNA curvature. It is to the users' advantage to have a choice for comparisons before jumping to conclusions. I have tried to build a bridge between 3DNA and Curves to make Curves users' life a bit more straightforward: find_pair have an -c option to generate input to Curves from a PDB file.

Since you are inquiring about "defining a local helix axis", the following message (slightly edited), which I communicated with an acquainted 3DNA user back on January 5, 2007, could be useful and/or more directly relevant.

> [2] my point is that the global axis is NOT displayed in the Raster3D mode.
> Nothing changes when I run frame_mol and then alc2img with the -g option.

[...]That 3DNA forum has been created for general discussions, so that the
community knows that 3DNA is still under active support and further
development.

3DNA certainly has more functionalities than described in our 2003 NAR
paper, and in reality, it is the details that count.

The simplest way to answer your 2nd question is by following an example
such as below:

Code: [Select]

find_pair bdl084.pdb stdout | analyze
rebuild bp_step.par bdl084.alc
rasmol -alchemy bdl084.alc
frame_mol -a -g -m -6,7 ref_frames.dat bdl084.alc bdl084_new.alc
rasmol -alchemy -noconnect bdl084_new.alc
alc2img -a -g -l bdl084_new.alc t.ps
gv t.ps
alc2img -r -a -g -l bdl084_new.alc t.r3d
render < t.r3d > t.avs
display t.avsThe global axis is displayed in t.avs, or t.ps with x-, y- etc labels.

Note: The above "recipe" works with 3DNA v1.5; One should use RasMol 2.6.x, not v2.7.x (a current version of Jmol is also fine) to display ALCHEMY format files properly; The command 'display' is from the ImageMagick package. See 3DNA 2008 NP paper for more examples.
[hr:611u016p][/hr:611u016p]
HTH,

Xiang-Jun

1481

General discussions (Q&As) / Re: Threading DNA

« on: June 22, 2009, 10:17:58 pm »

I wonder if it's possible using 3DNA to superpose a given(input) DNA sequence to the conformation of the DNA in the known protein-DNA complexes from the PDB database.

The short answer is no: 3DNA does not have such functionality built in. However, 3DNA is a tool set that could be used for such purpose. For an example, have a look http://www.ncbi.nlm.nih.gov/pubmed/19289051.

Xiang-Jun

1482

SCHNAaP/SCHNArP / Re: GLH_build() Memory Leak

« on: June 01, 2009, 07:00:27 pm »

I have fixed the memory leak bugs in file rebuild.c. I have also taken this opportunity to tidy up the code base a little bit. Please download the updated version and verify that the bugs are indeed gone.

Thanks,

Xiang-Jun

1483

General discussions (Q&As) / Re: How to make blocview -i work with pymol in batch mode.

« on: May 20, 2009, 07:03:35 pm »

Hi Mauricio,

Thanks for your note. In the current version of v2.0, the system call is as follows:

Code: [Select]

system("pymol -c $pmlfile 2&> x3dna_r3d_pymol.msg");i.e., with the -c option to run pymol in command-line mode. If I remember it correctly, I added this -c option following your suggestion (thanks!),  while you and Guohui helped verifying the recipes for the 2008 3DNA Nature Protocols paper.

I understand what -q means, and could add it to 3DNA v2.0 as you suggested (i.e., -qc). Does it really matter? What differences does it make: -c vs -qc? Could you have a try and report back?

While 3DNA is distributed in binary form for C programs (per Rutgers license policy), the Perl scripts are actually open source. It has been my intention to keep it that way so that users can follow the examples, make changes as needs arise, and create new tools.

Xiang-Jun

1484

General discussions (Q&As) / Please join me in my blog posts!

« on: May 17, 2009, 01:39:42 pm »

I have recently started to blog on topics I am interested in or feel worthwhile to take a note on. The posts necessarily reflect my personal, opinionated views, and you are welcome to make comments.

Xiang-Jun

1485

SCHNAaP/SCHNArP / Re: GLH_build() Memory Leak

« on: May 15, 2009, 11:24:59 pm »

Thanks for reporting the problems you experienced. Could you please be more specific, and better yet, providing me a reproducible example? I will certainly look into it to get it fixed ASAP.

Xiang-Jun

1486

General discussions (Q&As) / Re: Noobie question on building triplex DNA

« on: April 30, 2009, 10:50:49 pm »

Sorry for the noobie question, but is it possible to build a triplex DNA from scratch? If not, is it possible to take a .pdb file of a triplex DNA and mutate it to my own sequence? This would involve insertion of an extra residue in the loop region. I would later like to save the new coordinates in .pdb format for molecular dynamics studies.

This is good question: it clearly illustrates the need for a tool to manipulate/build/edit nucleic acid structures in 3-dimensional space. Ideally, it would be with an interactive GUI. Aren't there tools already available, commercial or free, that fit the bill?

As far as 3DNA goes, some of its 55 fiber models are triplexes. Starting from a PDB file of a triplex DNA and then mutating it to your specific sequence, or inserting an extra residue in the loop region,  are doable with 3DNA, along the same line as shown the thread "mutating DNA in DNA protein complex". Of course, the operations are not automated -- tedious and error-prone: you've got to know exactly what you want to do, and better have some programming skills to perform necessary geometric transformations.

Overall, 3DNA can do a lot, but there are many more things that can not be done easily or at all. It is a long journey from SCHNAaP/SCHNArP to 3DNA, and I have felt strongly for quite some time the need to move forward -- based on my 10+ years experience in creating and supporting SCHNAaP/SCHNArP/3DNA,  I have been building a new set of tools that can better meet the challenge of structure explosions. Among its many possible functionalities, it would make some of the requests in the forum, including this one, possible or much easier.

HTH,

Xiang-Jun

1487

General discussions (Q&As) / Re: how to minimize DNA built with fiber?

« on: April 22, 2009, 11:49:17 pm »

Could you be more specific? Which model did you use, and what's the sequence? The fiber models from 3DNA should have standard geometry and proper linkage. I am wondering how it could have broken bonds.

Xiang-Jun

1488

General discussions (Q&As) / Re: 3DNA version 2.0

« on: April 20, 2009, 11:33:53 pm »

Dear Ramon,

Thanks for bringing up the issue of downloading 3DNA v2.0. Dr. Olson is the proper contact for further instructions. Rutgers has imposed some licensing policy that is beyond my interest to get a hand on. Here is an excerpt, verbatim, of the message I sent to Dr. Olson regarding the release of 3DNA v2.0, after I uploaded the distribution tarball files for the most common operating systems (that I have access to) to the server in a password protected directory:

It is quite a while since our 3DNA NP paper was published. Over the
time I have received a couple of requests to download 3DNA v2.0
mentioned in our paper. Due to (complicated) license issues, I think
you are in a position to handle this issue. By getting more directly
involved, you would understand better what it means to maintain and
support a software.

As far as what's new with v2.0, it has already been mentioned in http://3dna.rutgers.edu/. To recap, v1.5 as currently available in http://rutchem.rutgers.edu/~xiangjun/3DNA/download.html was compiled 6 years ago that had never been updated. Of course, it still serves the majority of common cases very well, and I have used it as a test case of how robust my original implementation is (which is quite assuring, BTW). Of course, over the years, I have fixed bugs, added new features, and updated 3DNA to handle the remediated PDB files etc, which I have communicated with users on a case-by-case basis. The v2.0 contains an accumulation of all these refinements. More importantly, in writing the 3DNA Nature Protocols paper, I have added more command line options, new scripts and worked examples (NP_Recipes.tar.gz). Accurately reproducing these examples and understanding how each works, users would see why 3DNA is called 'a versatile, integrated software system for the analysis, rebuilding and visualization of three-dimensional nucleic-acid structures' . Aren't these sufficient reasons to upgrade?

Xiang-Jun

1489

General discussions (Q&As) / Re: how the cartesian coordinates transform to PDB format

« on: April 20, 2009, 11:14:46 pm »

First, thanks to Ramon for getting actively involved in answering other user's question. Over the years, it is my hope that 3DNA forum could turn into a virtual community where more people would participate in discussing issues related to nucleic acid structures. I am hoping others will follow your lead, and the 3DNA forum becomes more active.

Now to Si-Ya's question. 3DNA starts from a nucleic-acid containing structure in PDB format. Note specifically the coordinate section. As a more concrete example, have a look of the residue DT8 in chain A of entry 355d, as shown below:

Code: [Select]

ATOM    142  P    DT A   8       5.196  18.285   8.120  1.00 13.16           P 
ATOM    143  OP1  DT A   8       3.928  18.831   8.653  1.00 14.21           O 
ATOM    144  OP2  DT A   8       5.211  16.970   7.475  1.00 12.40           O 
ATOM    145  O5'  DT A   8       5.818  19.323   7.094  1.00 12.21           O 
ATOM    146  C5'  DT A   8       6.104  20.657   7.510  1.00 10.87           C 
ATOM    147  C4'  DT A   8       6.937  21.347   6.466  1.00  9.09           C 
ATOM    148  O4'  DT A   8       8.271  20.815   6.382  1.00  8.32           O 
ATOM    149  C3'  DT A   8       6.372  21.324   5.049  1.00  9.83           C 
ATOM    150  O3'  DT A   8       6.060  22.664   4.718  1.00 11.88           O 
ATOM    151  C2'  DT A   8       7.476  20.700   4.203  1.00  8.59           C 
ATOM    152  C1'  DT A   8       8.709  20.942   5.040  1.00  7.33           C 
ATOM    153  N1   DT A   8       9.786  19.985   4.858  1.00  7.74           N 
ATOM    154  C2   DT A   8      11.028  20.464   4.498  1.00  6.25           C 
ATOM    155  O2   DT A   8      11.253  21.654   4.285  1.00  7.74           O 
ATOM    156  N3   DT A   8      12.003  19.496   4.402  1.00  6.29           N 
ATOM    157  C4   DT A   8      11.852  18.139   4.631  1.00  5.16           C 
ATOM    158  O4   DT A   8      12.819  17.406   4.547  1.00  6.98           O 
ATOM    159  C5   DT A   8      10.502  17.708   4.979  1.00  5.39           C 
ATOM    160  C7   DT A   8      10.230  16.254   5.214  1.00  6.78           C 
ATOM    161  C6   DT A   8       9.556  18.638   5.074  1.00  5.19           C 
It is not just about the coordinates, but also about the naming convention of the base and backbone atoms. For example, for thymine, you have N1--C2--N3--C4--C5--C6 ring atoms, and O2 and O4 atoms attaching to C2 and C4.

As your example shows, it is clearly not in proper PDB format. It seems to be in xyz format. One might consider using 'babel' to convert it into PDB format. However, this converted version is not the one accepted by 3DNA, for reasons detailed in the above paragraph. I vaguely remember there is some tool to do proper conversion to PDB with correct atom names. Google it to see for yourself. For your specific purpose, I guess the 'simplest' way is to write a script to perform the conversion by taking into atom name convention into consideration.

HTH,

Xiang-Jun

1490

General discussions (Q&As) / Re: Find DNA-protien contacts using 3DNA

« on: April 18, 2009, 01:45:09 pm »

Is there a direct option/function to find amino acid atoms around base pairs, i.e. protein contacts with DNA/RNA? Or can you suggest a way to combine various 3DNA functions to approach this?

No, there is no such as an option in 3DNA to find contacts of amino-acids with nucleotides, as the "-w" option for waters.

In the same paper, I also mentioned:

The interactions with users from different backgrounds have given us the incentive to adapt the programs for further applications in related fields, for example, RNA structure-motif identification and alignment, structural analysis of DNA–protein complexes and modeling RNA folds. The reference-frame-based description of three-dimensional spatial geometry makes the methodology and algorithms in 3DNA directly applicable to these problems and treats them in a rigorous and consistent fashion

Specifically, for one of my current research projects, I have written a program named SNAP (Structure of Nucleic Acids and Protein) which has this functionality, among other things. However, SNAP is not part of 3DNA: I am currently beginning to write a manuscript on this topic, and will make it available in due time.

Of course, if you check the literature, there are many ways to find the contacts between aa with nt: e.g., Kono/Sarai, Pabo + Siggers/Honig, and Luscombe/Thornton.

Xiang-Jun

[hr:1je7sv3u][/hr:1je7sv3u]
PS. As a side note, in 3DNA Nucleic Acids Research, 2003 (Vol. 31, No. 17, 5108--5121), I mentioned the following:

Since the six base pair parameters uniquely define the relative position and orientation of two bases, they can be used to reconstruct the base pair. Moreover, the parameters provide a simple mechanism for classification of structures (55) and database searching (X.-J. Lu, Y. Xin and W.K. Olson, unpublished data).

1491

General discussions (Q&As) / Re: Calculation of Jtwist, Jroll and Jslide for junctions

« on: April 01, 2009, 09:07:51 pm »

Thanks for the reference. While 3DNA is not directly applicable for calculating such parameters, it could help in some ways, e.g., to identify the double helical regions and get the helical axes etc.

Of course, you could always check with the original authors for further technical details (e.g., the program to reproduce their results). If you want to implement their method yourself and would like to share, I am willing to help in a sensible way.

Xiang-Jun

1492

General discussions (Q&As) / Re: Calculation of Jtwist, Jroll and Jslide for junctions

« on: March 31, 2009, 11:25:50 pm »

How to calculate Jtwist, Jslide and Jroll using 3DNA software for junction structure?

I am not familiar with the names Jtwist, Jslide and Jroll, and they are not calculated by 3DNA.

Would you mind sharing more details, e.g., providing a reference?

Xiang-Jun

1493

SCHNAaP/SCHNArP / Re: SCHNArP: Global Parameters

« on: March 20, 2009, 10:58:05 pm »

Well, once you try to get into details on how things actually work in SCHNAarP, or any software tools in that matter,  you will surely have lots of questions. Getting the software compiled and run is just the beginning.

Overall, SCHNAarP was produced 10+ year ago, and it is now superseded by 3DNA (v2.0). That does not mean the underlying algorithms are out of date. Just on the contrary, the mathematics is valid and solid, and it forms one of the foundations of 3DNA. Yet by design, the reference frames in CEHS and SCHNAaP only apply to double helical DNA/RNA structures (with Waton-Crick bps). As a special note, stretch for a Watson-Crick base-pair is ~5.4 A instead of 0 A, as would be expected, and from other analysis programs.

Now for your specific questions:

There is no special documentation to the file format on Sequence-GLH parameter file. It would be self-explanatory by following an example. For your case, first enter #2 for "Use GLOBAL helical parameters." Then I enter #1 for "Uniform regular helix" and you will get an output file "GLH_seq.dat". Examine it and post back here what you find. And have a look of 1bna.glh following schnaap.

Building RNA structure from a set of SCHNAaP global parameters would be practically meaningless (see above). With the source code in hand and once you get to the bottom of it, you could borrow the idea to apply to your specific applications. This does take time and efforts -- it is not just about the C code, but more about the underlying mathematics and the nucleic acid structure problems being addressed.

HTH,

Xiang-Jun

1494

SCHNAaP/SCHNArP / Re: SCHNArP/SCHNAaP c source code

« on: March 18, 2009, 11:59:27 pm »

All bug fixes / changes have been incorporated into the code base from the download site. Thanks to 'clarebonk' for reporting and testing!

As always, bug reports and comments are gratefully received.

Xiang-Jun

1495

SCHNAaP/SCHNArP / Re: SCHNArP/SCHNAaP c source code

« on: March 17, 2009, 10:18:33 pm »

Very strange. It should work -:)

What OS and compiler are you using? Is it 'gcc'? Did you change all four NR functions? For verification purpose, please send me your finished 'cmn_fncs.c' file by email.

Xiang-Jun

1496

General discussions (Q&As) / Re: Building a DNA superhelix

« on: March 04, 2009, 11:16:35 pm »

Thanks for the info. Just to be precise, the link you gave is broken: it should also include the ending characters in the URL. For the benefit of other viewers, here is the correct one with title of the paper as link text: Geometry of the Nucleosomal DNA Superhelix by Thomas C. Bishop. (We all make simple mistakes, and I hope you won't mind my pointing your minor typo out  :roll:. I am hoping you and others will do the same to any 3DNA-related issues. Every bug report or correction etc would be gratefully received!

I have no problem in accessing this paper. I think it is a nice piece of work. It is actually in the 3DNA citation list. I should have read the abstract before, but did not get into the main text. To be realistic, though, I do not think I have that much spare time right now to understand and implement the algorithm to fit your purpose -- it won't be a trivial take, especially to get it done efficiently and robustly. You could contact the author directly for further details. Maybe we could make an arrangement to incorporate Dr. Bishop's method into 3DNA (with proper attributions, of course).

Thanks for your cooperation, and hopefully more 3DNA users would follow your example.

Xiang-Jun

1497

General discussions (Q&As) / Re: Building a DNA superhelix

« on: March 03, 2009, 11:46:11 pm »

Dear Miguel,

Thanks for posting your DNA super-helix building question here. Currently, 3DNA does not yet have the functionality. As you noticed, Arvind posted a related question, i.e., analyzing a DNA super-helix to give its pitch and radius.

Essentially, this is the same issue looked from different perspectives, in a sense similar to SCHNAaP/SCHNArP and analyze/rebuild for DNA/RNA duplex. Any rigorous treatment should be mathematically reversible, i.e., given a super-helix, one can derive a set of pitch/radius parameters; conversely, from this same set of parameters, the original super-helix should be rebuilt without loss of information. I would like to hear more on related issues in the literature. Any clues?

Xiang-Jun

1498

General discussions (Q&As) / Re: Base Pair Step Parameters

« on: February 21, 2009, 01:05:55 pm »

Hi Rodrigo,

Your case is yet another example of the importance to be specific when asking questions and in discussions.

From your attached 3DNA output file, the H-bonds are normal for what would be expected for Watson-Crick pairs:

Code: [Select]

****************************************************************************
Detailed H-bond information: atom-name pair and length [ON]
   1 G-----C  [3]  O6 - N4  2.86  N1 - N3  2.90  N2 - O2  2.81
   2 A-----U  [2]  N6 - O4  3.03  N1 - N3  2.90
   3 A-----U  [2]  N6 - O4  2.98  N1 - N3  2.91
   4 A-----U  [2]  N6 - O4  2.94  N1 - N3  2.95
   5 A-----U  [2]  N6 - O4  2.89  N1 - N3  2.96
   6 G-----C  [3]  O6 - N4  2.95  N1 - N3  2.93  N2 - O2  2.85
   7 A-----U  [2]  N6 - O4  3.05  N1 - N3  2.92
   8 A-----U  [2]  N6 - O4  3.03  N1 - N3  2.94
   9 A-----U  [2]  N6 - O4  2.96  N1 - N3  2.92
  10 G-----C  [3]  O6 - N4  2.91  N1 - N3  2.91  N2 - O2  2.80
  11 A-----U  [2]  N6 - O4  3.00  N1 - N3  2.89
  12 A-----U  [2]  N6 - O4  2.95  N1 - N3  2.96
****************************************************************************

Naturally, the six base-pair parameters have mean and std values in ranges reported in the literature:

Code: [Select]

****************************************************************************
Local base-pair parameters
     bp        Shear    Stretch   Stagger    Buckle  Propeller  Opening
    1 G-C      -0.23     -0.18     -0.63    -20.29     -2.28     -0.66
    2 A-U       0.09     -0.04     -0.48    -13.64    -10.06      4.26
    3 A-U       0.06     -0.06     -0.53    -11.24    -12.42      2.82
    4 A-U       0.09     -0.03     -0.62    -13.46    -13.99      0.86
    5 A-U       0.18      0.01     -0.56    -15.56    -12.75     -1.66
    6 G-C      -0.05     -0.03     -0.69    -25.26    -19.76      2.36
    7 A-U       0.06      0.01     -0.35     -7.49     -7.94      4.36
    8 A-U       0.14     -0.00     -0.49     -8.36    -11.15      3.57
    9 A-U       0.09     -0.04     -0.34     -6.42    -10.45      1.32
   10 G-C      -0.11     -0.13     -0.33     -9.45    -12.75      1.08
   11 A-U       0.08     -0.08     -0.36     -9.60    -12.41      3.48
   12 A-U       0.16     -0.09     -0.35     -1.29    -11.67      0.95
          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      ave.      0.05     -0.06     -0.48    -11.84    -11.47      1.89
      s.d.      0.12      0.06      0.13      6.42      4.05      1.91
****************************************************************************
Thus, from what I have in hand, nothing appears (to me) really unusual here. Of course, it would also be helpful to analyze your structure with other programs, e.g., Curves, to see what you get.

HTH,

Xiang-Jun

1499

General discussions (Q&As) / Re: Base Pair Step Parameters

« on: February 17, 2009, 10:09:53 pm »

I´m analyzing an double helix RNA with a modified nucleotide. The effect of the modification pushes one of the bases but doesnt affect complementary one. stragger, proppeler and Buckle are very affected. Since it affects only one base of the base pair I dont know if it is correct two analyze roll and tilt. May be its a wrong average of a non-averageble structure.

Again, it would make life easier for me and other viewers of this forum to understand exactly what you mean if you have provided more details,  i. e., the PDB file containing the fragment you are having trouble with and corresponding 3DNA output. Remember to help as much as possible to others so they can help you!

In a general sense, I believe I know what you mean, and I am glad that you bring up this issue. There is still confusions in the literature regarding analysis of non-Watson-Crick base-pairs and their associated helical regions, which are prevalent in RNA structures. Without going into specific details, I would suggest that you leave out each such pair and its two associated steps from averaging with other normal WC pairs/steps. A comparison is meaningful only if its parts are comparable, which is not the case for a WC vs non-WC pairs.

However, such non-WC pairs should not be ignored  -- instead, they should be treated separately and with more care. 3DNA can help in pursuing such cases further, especially in the context of the structural database (NDB/PDB), with its various visualization and analysis tools.

HTH,

Xiang-Jun

1500

General discussions (Q&As) / Re: stability of rna structure

« on: February 11, 2009, 08:39:27 pm »

Hi,

You have an interesting RNA structure. In agreement with you, I do not think that you can pick up "the best based on structural parameters alone". In addition to "give odd values" of the structural parameters, what would be your reference for your comparison? 3DNA does not fit the bill here. Maybe you could get more helpful feedbacks from mailing lists such as GROMACS, AMBER etc packages.

From a pure structural point of view, given a cluster of similar structures, one could perform all-vs-all RMSD fitting to find the "centroid" (i.e., the one with minimum overall RMSD against others) as a representative.

HTH,

Xiang-Jun