Messages

1151

MD simulations / Re: Stacking of two neighboring non-base-paired bases

« on: February 19, 2013, 10:03:16 pm »

Reviewing the thread, you would recall that find_pair is not used for your special case. The input file is prepared manually, and then fed to analyze to get the result. It is more likely something is wrong in your for-loop than a bug in 3DNA. If you understand the protocol as we went through, then try to write a for loop that repeats only once to confirm its correctness.

As always, a reproducible example would have made things clear.

Xiang-Jun

1152

General discussions (Q&As) / Re: Building chromosome with fiber

« on: February 18, 2013, 12:07:40 pm »

Hi Nikolay,

Thanks for providing further info. I tried to build a fiber-based structure with 48M bps on my new iMac with 32GM RAM. The fiber program did not exist with the memory allocation error as you saw, but it seems never to finish!

So fiber appears not up to the job, even though in principle it should. Note also that with 48M bps, the PDB format is certainly out of the question. If you really want to go for it, you may try smaller fragments (with the -xml option for the PDBML format) that can be built with fiber, and then assemble them into the whole 48M bps chromosome.

Xiang-Jun

1153

General discussions (Q&As) / Re: Building chromosome with fiber

« on: February 18, 2013, 10:18:38 am »

The error message you reported is related to memory allocation. It is likely that your computer is not up to the task. What are the settings of your computer? How much RAM do you have?

How did you run the fiber program specifically? How many bps does the chromosome have?

If you could provide such details, others may help you identify where the problem is, and possibly come up with a solution.

Xiang-Jun

1154

MD simulations / Re: Stacking of two neighboring non-base-paired bases

« on: February 17, 2013, 11:32:21 am »

This is where programming (scripting) skills come into play. You do not have to do this manually for each of your 25k+ frames, once you have written a purpose-speicifc script to automate the process. You may find it helpful to have a look of the Ruby script x3dna_ensemble and corresponding files under folder lib/.

Alternatively, Curves+ or other tools may better suit your needs.

Xiang-Jun

1155

General discussions (Q&As) / Re: B-Factor values from PDB in output files?

« on: February 13, 2013, 06:24:58 pm »

Thanks for your feedback -- that helps me in making decisions. I could add the new functionality to 3DNA's "analyze" program, or more sensibly, to the new RNA-focused component (DSSR) I am currently working on. I aim to get DSSR (alpha test version) out by the end of the month, or even possibly by next week.

NMR structures in PDB seem to have the b-factor columns filled. Calculating the average values of b-factors regardless structure type would make coding simple and consistent.

Xiang-Jun

1156

General discussions (Q&As) / Re: B-Factor values from PDB in output files?

« on: February 13, 2013, 01:34:02 pm »

Thanks for your feedback.

Just to be sure: by ".outp", you mean the option "find_pair -p"-generated file "allpairs.ana" which is then fed into "analyze"? How about the default setting? i.e., the ".out" file?

I will think more about this request, and I may come up with something. If I decide to go for it, I'd output the B-factor average/occupancy for any input structure, not just crystal structures. Moreover, there would be a new command line option for such info, which is OFF by default to be compatible with previous 3DNA releases.

To help me help you in this endeavor, could you provide (at least) a concrete example with the values you want?

Xiang-Jun

1157

General discussions (Q&As) / Re: pdf files to use with spdbv and Curves+

« on: February 13, 2013, 12:11:28 pm »

It certainly would help if you've included specifics as to how w3DNA-generated PDB files cannot be used with spdbv or Curves+. I guess the issue is due to PDB v2.x atom names used by w3DNA which is still not updated to v2.1 yet. In the current 3DNA release, PDB v3.x atom names are used by default, which should make "spdbv or Curves+" happy.

Please download and install 3DNA v2.1, and report back if that helps. See also the thread "Missing atoms".

HTH,

Xiang-Jun

1158

General discussions (Q&As) / Re: B-Factor values from PDB in output files?

« on: February 12, 2013, 08:48:25 pm »

The B-factor values are currently not included in 3DNA output file(s), and you are the first user to request such info. Where do you think is the appropriate place to put them? And in what format? How about atom occupancy?

Xiang-Jun

1159

General discussions (Q&As) / Re: How to include dotted line along helical axis in figure?

« on: February 04, 2013, 06:28:28 pm »

Could you be specific on what you mean by adding "the helical axis as a dotted line in the 3dna-generated figures"? As always, a concrete example would help clarify your point.

Fig. 3 (recipe #2) of the 3DNA NP08 paper shows a dotted line contacting bp centers, but I sense that's not what you want. Does the thread "defining a local helix axis" help?

Xiang-Jun

1160

MD simulations / Re: Stacking of two neighboring non-base-paired bases

« on: January 28, 2013, 11:07:30 am »

Upon your request, I've removed "the image of the whole structure". I understand your concern, and good luck with your manuscript!

Xiang-Jun

1161

MD simulations / Re: Stacking of two neighboring non-base-paired bases

« on: January 27, 2013, 11:22:07 pm »

Thanks for providing a PDB file which helps clarify the issue.

The find_pair program is working as designed. For your RNA structure, it identifies two helical regions (see the attached figure below --- [Note added on 2013-01-28: image deleted upon request]):

Code: [Select]

pdb.00001
pdb.out
    2         # duplex
   30         # number of base-pairs
    1    1    # explicit bp numbering/hetero atoms
   31    8  0 #    1 | ....>-:..31_:[.RG]G-----C[.RC]:...8_:-<....  0.23  0.08 16.57  8.85 -3.78
   32    7  0 #    2 | ....>-:..32_:[.RC]C-----G[.RG]:...7_:-<....  0.50  0.23 11.02  8.87 -3.49
   33    6  0 #    3 | ....>-:..33_:[.RA]A-----U[.RU]:...6_:-<....  0.15  0.03  4.87  8.77 -4.55
   34    5  0 #    4 | ....>-:..34_:[.RA]A-----U[.RU]:...5_:-<....  0.14  0.09  6.89  8.85 -4.32
   35    4  0 #    5 | ....>-:..35_:[.RG]G-----C[.RC]:...4_:-<....  0.84  0.37  4.54  8.71 -3.19
   36    3  0 #    6 | ....>-:..36_:[.RC]C-----G[.RG]:...3_:-<....  0.84  0.79 15.85  8.75 -1.79
   37    2  0 #    7 | ....>-:..37_:[.RU]U-**--G[.RG]:...2_:-<....  3.01  0.85 17.27  8.99  2.57
   38   23  0 #    8 | ....>-:..38_:[.RG]G-----C[.RC]:..23_:-<....  0.45  0.30 12.77  8.96 -3.31
   39   22  0 #    9 | ....>-:..39_:[.RG]G-----C[.RC]:..22_:-<....  0.50  0.33 33.29  8.85 -2.19
   42   40  0 #   10 | ....>-:..42_:[.RA]A-**+-G[.RG]:..40_:-<....  7.23  1.99 47.35  7.57 10.57
   43   61  0 #   11 | ....>-:..43_:[.RA]A-**--G[.RG]:..61_:-<....  1.60  0.88 20.41 10.70  1.38
   44   60  0 #   12 | ....>-:..44_:[.RC]C-----G[.RG]:..60_:-<....  0.34  0.04 24.77  8.89 -3.35
   45   59  0 #   13 | ....>-:..45_:[.RA]A-----U[.RU]:..59_:-<....  0.13  0.06 17.54  8.79 -3.87
   46   58  0 #   14 | ....>-:..46_:[.RU]U-----A[.RA]:..58_:-<....  0.11  0.07 12.05  8.78 -4.15
   47   57  0 #   15 | ....>-:..47_:[.RU]U-----A[.RA]:..57_:-<....  0.36  0.34 10.79  8.72 -3.42
   48   55  0 #   16 | ....>-:..48_:[.RC]C-----G[.RG]:..55_:-<....  0.40  0.02 28.47  8.84 -3.15
   49   54  0 #   17 | ....>-:..49_:[.RC]C-----G[.RG]:..54_:-<....  0.53  0.38 22.85  8.85 -2.56
   50   53  9 #   18 x ....>-:..50_:[.RG]G-**--A[.RA]:..53_:-<....  9.64  0.99 16.65  8.51 11.45
   10   73  0 #   19 | ....>-:..10_:[RG5]g-----c[RC3]:..73_:-<....  0.37  0.20 12.89  8.98 -3.58
   11   72  0 #   20 | ....>-:..11_:[.RG]G-----C[.RC]:..72_:-<....  0.97  0.26 20.72  8.78 -2.48
   12   71  0 #   21 | ....>-:..12_:[.RU]U-----A[.RA]:..71_:-<....  0.17  0.04 22.04  9.00 -3.65
   13   70  0 #   22 | ....>-:..13_:[.RC]C-----G[.RG]:..70_:-<....  0.46  0.18 17.81  9.02 -3.30
   14   69  0 #   23 | ....>-:..14_:[.RC]C-----G[.RG]:..69_:-<....  0.52  0.46 10.86  9.01 -3.01
   15   68  0 #   24 | ....>-:..15_:[.RG]G-----C[.RC]:..68_:-<....  0.33  0.29 12.02  8.95 -3.50
   16   67  0 #   25 | ....>-:..16_:[.RC]C-----G[.RG]:..67_:-<....  0.36  0.30 11.63  8.96 -3.45
   17   66  0 #   26 | ....>-:..17_:[.RA]A-----U[.RU]:..66_:-<....  0.18  0.00  4.99  8.79 -4.57
   18   30  0 #   27 | ....>-:..18_:[.RG]G-----C[.RC]:..30_:-<....  0.35  0.07 24.12  8.82 -3.30
   19   29  0 #   28 | ....>-:..19_:[.RC]C-----G[.RG]:..29_:-<....  0.49  0.20 24.09  8.88 -2.91
   20   28  0 #   29 | ....>-:..20_:[.RC]C-----G[.RG]:..28_:-<....  0.73  0.69 13.00  9.01 -2.25
   21   26  0 #   30 | ....>-:..21_:[.RU]U-**+-G[.RG]:..26_:-<....  1.30  0.34 16.05  9.43 -0.22
##### Base-pair criteria used:     4.00     0.00    15.00     2.50    65.00     4.50     7.80 [ O N]
##### 5 non-Watson-Crick base-pairs, and 2 helices (0 isolated bps)
##### Helix #1 (18): 1 - 18  ***broken O3'[i] to P[i+1] linkage***
##### Helix #2 (12): 19 - 30  ***broken O3'[i] to P[i+1] linkage***
G1 and U24 are not paired according to the default criteria. Indeed, as shown below and as mentioned in your initial post, G1 and U24 are stacking instead of pairing.

I want to analyze two bases in a ribozyme, G1 and U24, neither of which are in any base pairs. I want to track the sliding interaction of these two bases across the total length of my trajectory, so that I can correlate the sliding to various other distances and angles that I measure across the length of the trajectory as well. I can see how one can extract the slide parameter for a base pair step, but I don't understand if and how one can extract the slide parameter for two unpaired bases.

If you insist on finding the relative geometry of G1 to U24, you could play the following trick:

Code: [Select]

find_pair -s pdb.00001 pdb.00001.datThe -s option means to output a list of all nucleotides in the given PDB file. Since you are interested in only G1 vs U24, you can manually edit the above output file pdb.00001.dat to have the following content:

Code: [Select]

pdb.00001
pdb.outs
    1      # single helix
    2      # number of bases
    1    1 # explicit bp numbering/hetero atoms
    1      #     1 ....>-:...1_:[RG5]g
   24      #    24 ....>-:..24_:[.RU]U
Running "analyze pdb.00001.dat", you get in file pdb.outs the following section:

Code: [Select]

****************************************************************************
Local base step parameters
    step       Shift     Slide      Rise      Tilt      Roll     Twist
   1  g/U      -1.95     -2.62      1.18    -29.34    146.33     78.47
****************************************************************************
Try a few frames (snapshots) along your MD simulation trajectories to decide for yourself if that make sense -- certainly this usage is beyond 3DNA's 'normal' application range.

Xiang-Jun

1162

MD simulations / Re: Stacking of two neighboring non-base-paired bases

« on: January 27, 2013, 09:00:49 am »

Thank for providing further details. However, for your example to be reproducible, you need to provide a PDB file -- that's the starting point of all the following steps.

Basically, you need to run find_pair first to get the pairing info in a file (as the one you attached); the file may need to be modified manually if some desired pair is missing, as is most likely in your case. Then one runs x3dna_ensemble analyze to get the various 3DNA parameters, followed by x3dna_ensemble extract to pick up the parameters one is interested in. The first step in analyzing an ensemble (MD or NMR), i.e, preparing a correct bp file, is identical to that for a single structure, and it is the most crucial.

In the coming 3DNA JoVE paper, there is a protocol to illustrate the whole procedure. In the meantime, you may want to try the examples distributed with 3DNA to get familiar on how to run x3dna_ensemble.

Xiang-Jun

1163

MD simulations / Re: Stacking of two neighboring non-base-paired bases

« on: January 26, 2013, 11:36:40 pm »

Welcome to join the 3DNA-user community!

Please note that the x3dna_md.rb and extract_par.rb pair has been replaced by the x3dna_ensemble script, which consolidates ensemble-processing functionality in 3DNA v2.1 under one umbrella. Please update your 3DNA installation to the latest version.

Regarding your specific question, could you provide a reproducible example? A single frame from your MD simulations can be used to illustrate your point. Show step-by-step what you want to achieve, what's missing from 3DNA, and we can start from there.

Xiang-Jun

1164

General discussions (Q&As) / Re: Analyzing some pdb files of bent DNA

« on: January 15, 2013, 12:15:40 pm »

The helix break is controlled by a parameter callled helix_break in file $X3DNA/config/misc_3dna.par:

Code: [Select]

#   distance criterion for helix break
<helix_break>7.5</helix_break>
It has a default value of 7.5 Å. Reset it to a larger value, e.g. 8.0 Å (see FAQs), will do the trick for all your cases.

HTH,

Xiang-Jun

1165

General discussions (Q&As) / Re: Analyzing some pdb files of bent DNA

« on: January 14, 2013, 04:22:19 pm »

Thanks for using w3DNA, a web-service hosted and supported by the Olson laboratory at Rutgers University. w3DNA aims to make commonly/routinely used 3DNA functionality easily available. To taking full advantage of 3DNA, the standard command-line version is the way to go. Also, w3DNA is still using 3DNA v2.0, while the the command-line version is v2.1.

The issue you experienced for PDB entries 2o8b and 2o8c, i.e., some step parameters are designated as "----", is due to the kink in the two structures. Thus, find_pair now takes each DNA as two helical regions instead of a continuous helix, see below:

find_pair 2o8b.pdb 2o8b.inp
# contents of 2o8b.inp
2o8b.pdb
2o8b.out
    2         # duplex
   15         # number of base-pairs
    1    1    # explicit bp numbering/hetero atoms
    1   30  0 #    1 | ....>E:...1_:[.DG]G-----C[.DC]:..30_:F<....  1.14  0.36 14.09  9.30 -0.44
    2   29  0 #    2 | ....>E:...2_:[.DA]A-----T[.DT]:..29_:F<....  0.89  0.82 26.82  9.09 -1.13
    3   28  0 #    3 | ....>E:...3_:[.DA]A-----T[.DT]:..28_:F<....  0.23  0.03 18.38  9.24 -3.78
    4   27  0 #    4 | ....>E:...4_:[.DC]C-----G[.DG]:..27_:F<....  0.78  0.19  8.59  8.93 -3.40
    5   26  0 #    5 | ....>E:...5_:[.DC]C-----G[.DG]:..26_:F<....  0.56  0.29 17.33  9.24 -3.00
    6   25  0 #    6 | ....>E:...6_:[.DG]G-----C[.DC]:..25_:F<....  0.29  0.27 19.28  9.01 -3.20
    7   24  9 #    7 x ....>E:...7_:[.DC]C-----G[.DG]:..24_:F<....  0.22  0.01 24.18  9.04 -3.55
    8   23  0 #    8 | ....>E:...8_:[.DG]G-**--T[.DT]:..23_:F<....  5.22  0.31 43.28  9.87  7.00
    9   22  0 #    9 | ....>E:...9_:[.DC]C-----G[.DG]:..22_:F<....  0.44  0.33 20.98  8.84 -2.85
   10   21  0 #   10 | ....>E:..10_:[.DG]G-----C[.DC]:..21_:F<....  0.41  0.39 10.80  9.05 -3.28
   11   20  0 #   11 | ....>E:..11_:[.DC]C-----G[.DG]:..20_:F<....  0.26  0.01 12.86  9.06 -4.07
   12   19  0 #   12 | ....>E:..12_:[.DT]T-----A[.DA]:..19_:F<....  0.85  0.47 11.88  8.95 -2.62
   13   18  0 #   13 | ....>E:..13_:[.DA]A-----T[.DT]:..18_:F<....  0.53  0.18  9.30  8.85 -3.65
   14   17  0 #   14 | ....>E:..14_:[.DG]G-----C[.DC]:..17_:F<....  1.17  0.94 21.28  8.97 -0.89
   15   16  0 #   15 | ....>E:..15_:[.DG]G-----C[.DC]:..16_:F<....  1.17  0.05 37.85  8.66 -1.83
##### Base-pair criteria used:     4.00     0.00    15.00     2.50    65.00     4.50     7.50 [ O N]
##### 1 non-Watson-Crick base-pair, and 2 helices (0 isolated bps)
##### Helix #1 (7 ): 1 - 7
##### Helix #2 (8 ): 8 - 15

You can fix the problem by either changing 9 for bp #7 to 0, or running analyze with the -c option:

Code: [Select]

analyze -c 2o8b.inp
In 3DNA, lower case a/c/g/t/u is used for modified bases of A/C/G/T/U respectively. For example, in 2o8c, 6OG is assigned as g.

HTH,

Xiang-Jun

1166

General discussions (Q&As) / Re: question on +/- signs of local bp parameters

« on: January 13, 2013, 11:21:29 pm »

Hi Jane,

I do not quite get what you mean for your follow up question.

Regarding the local reference frame, please have a look of the 2003 3DNA paper, especially Figure 2 (below):


and Figure 1 of the base reference frame paper:


Specifically, the x-axis points from the minor-groove side to the major groove side. For the aligned A and T bases, it's horizontal, pointing to the right. It's a good exercise for you to work out where the x-axes for DT-104 and DA-105 point to. Then you should be able to figure out why one opening is positive, and another is negative.

Note that in 3DNA, the reference frame is defined purely based on the base atoms, not taking consideration of the syn/anti chi torsion angle. See my post "The chi (χ) torsion angle characterizes base/sugar relative orientation".

Xiang-Jun

1167

General discussions (Q&As) / Re: question on +/- signs of local bp parameters

« on: January 11, 2013, 03:57:13 pm »

Thanks for providing an example that helps to illustrate the 'asymmetry' of the A-T vs T-A Hoogsteen pair when put in a structural context. The difference in opening can be understood in the same way as shear in G-U vs U-G wobble pair. See the figure below where the A and T on one strand are aligned on their base reference frames; note the different orientations of T and A.

In 3DNA, the Hoogsteen pair is of the M+N type; thus if the bases are swapped to N+M, all bp parameters change signs.

Code: [Select]

    3 T+A      -0.52      3.62      0.46     -6.35     10.24    -67.69
    4 A+T       0.60     -3.57     -0.56      6.51    -11.43     67.59
3DNA is unique in rigorously quantifying such differences using the six rigid-body bp parameters.

HTH,

Xiang-Jun

1168

General discussions (Q&As) / Re: question on +/- signs of local bp parameters

« on: January 11, 2013, 01:51:54 pm »

Thanks for using 3DNA and posting your questions on the Forum.

For B-DNA structures, some base-pair parameters (e.g., shear, opening) are normally quite small, so the difference between +/- values are not that obvious. For some extreme cases, however, the meaning in +/- is immediately obvious. For example, the famous U-G & G-U wobble pairs are distinguished by +/- shear -- see my post "Difference in shear of neighboring base pairs affects twist angle".

As for opening, if you could provide some concrete examples, I will help you work them through to see the difference.

HTH,

Xiang-Jun
 

1169

w3DNA -- web interface to 3DNA / Re: Structural quality

« on: January 11, 2013, 12:26:12 pm »

Hi Damien,

Please note that 3DNA works 'mechanically' -- it calculates a set of parameters for your input structure, without checking if it is reasonable at all. However, since an erroneous structure often gives some bizarre parameters, 3DNA can give a careful user some hint on possible issues.

A quick check using Jmol (or RasMol) shows one region of your model is displayed differently from the rest, signaling some problems. For a detailed report regarding the quality of your model, please try MolProbity (http://molprobity.biochem.duke.edu/).

Xiang-Jun

1170

General discussions (Q&As) / Re: Illegal instruction 4

« on: January 10, 2013, 12:02:14 pm »

Hi Susana,

Thanks for reporting the "Illegal instruction" error message when you ran 3DNA on your Mac OS X Lion 10.7.5 machine. It was due to a backward compatibility issue for the 3DNA binaries I compiled on a newer 10.8.x (Mountain Lion) Mac. I've recompiled 3DNA on a Mac OS X 10.6 (Snow Leopard) and now it should work for you; please download the 3DNA 2013jan10 version from the download page and report back how it goes.

I am using a MacBook Air as the primary machine to develop 3DNA, so a native Mac OS X version of the software should always be available.

HTH,

Xiang-Jun

1171

General discussions (Q&As) / Re: How to build a dsRNA helix with non-Watson-Crick base-pairs

« on: January 09, 2013, 08:33:43 pm »

3DNA does not provide an automatic way to build a dsRNA with non-Watson-Crick base pairs, e.g., the G-U wobble pair. Nevertheless, it has the components that can could possibly be combined to get the job done, on a case-by-case basis. If you are specific about what you want to achieve, our discussions can be more concerete. Alternatively, you may try software tools developed from the laboratories of Eric Westhof or François Major, among other possible choices.

Xiang-Jun

1172

w3DNA -- web interface to 3DNA / Re: misshapen base ring

« on: January 08, 2013, 01:19:16 pm »

Hi Damien,

I've updated 3DNA v2.1 to 2013jan08, with an improved algorithm for nucleotide identification. Specifically, in its default setting, DT8 on chain F is no longer recognized as a nucleotide. For the record, in my yesterday's response with an attached image, I observed the missing pair issue via Jmol using the DT8 as a test case. However, I did not check carefully to notice that previous versions of 3DNA, by virtue of very general distance cutoffs, actually take it as a nucleotide!

Note that the w3DNA web-server hosted at Rutgers is currently not yet updated to the latest version. So your best bet is to download the standard command-line version of 3DNA; it is the most efficient and convenient way to get your job done.

HTH,

Xiang-Jun

1173

w3DNA -- web interface to 3DNA / Re: misshapen base ring

« on: January 07, 2013, 01:34:38 pm »

Hi Damien,

Thanks for your follow-up. Yes, you can certainly use 3DNA to detect anomaly in a structure; that happened a (long) while ago at the NDB.

Thanks for pointing out the "bad" pair involving DT8 on chain F. The issue is due to the very 'generous' distance criteria used, which work well for 'reasonable' cases but apparently fail for your severely distorted structure. I'll fix the issue and update 3DNA shortly.

Xiang-Jun

1174

w3DNA -- web interface to 3DNA / Re: misshapen base ring

« on: January 07, 2013, 11:55:26 am »

Hi Damien,

Thanks for providing a sample PDB file that illustrates the problem you are facing; it helped me to identify the issue.

As shown by the attached image for DT8 on chain F, the base is distorted beyond recognition. For example, C5--C6 distance is only 0.42 Å -- far too short for a covalent bond, whereas N1--C6 = 1.97 Å and C4--C5 = 2.16 Å are far too long. So 3DNA won't recognize it as a DNA base T at all. Same issues exist for the other three nucleotides.

It does not makes much sense to change 3DNA to accommodate such erroneous cases; rather, the mistakes should be fixed in any tool you used to generate this structure in the first place.

HTH,

Xiang-Jun
 

1175

w3DNA -- web interface to 3DNA / Re: misshapen base ring

« on: January 07, 2013, 10:45:36 am »

Could you post (attach) an example structure to make your point concreate?

Xiang-Jun